Antibody-drug conjugates (ADCs) are advancing cancer treatment, but their success depends on early safety insights and informed toxicology strategies. Xiaoxiao Wang, Yan Zhang, Junli Zhao, WuXi AppTec explains.
Shutterstock - agustin.photo
Antibody-drug conjugates are changing what is possible in oncology and the treatment of other complex conditions by pairing targeted delivery with highly potent cytotoxic activity. For patients, that promise is meaningful: more precise therapies, new options for difficult-to-treat cancers, and the potential to extend benefit while limiting avoidable harm. But realising that promise depends on an early and nuanced understanding of safety risks. In ADC development, toxicology strategy must go beyond detecting risk to understanding where that risk comes from and carefully considering how it should guide clinical decisions.
Why ADC toxicology is different
In conventional drug development, nonclinical safety assessment is designed to characterise the relationship between pharmacologic activity, systemic exposure, and adverse effects. For many small molecules, for instance, that may distribute widely and cause systemic toxicities, toxicology programmes focus on defining dose-dependent target organ effects, exposure-toxicity relationships, and an optimal therapeutic window. For conventional monoclonal antibodies, adverse effects are more commonly associated with excessive on-target pharmacology, underscoring the importance of target tissue distribution, pharmacologic species relevance, and mechanism-based safety monitoring in risk assessment.
ADCs are different. Their unique structure and function, distinct from other modalities, require a more integrated toxicology strategy. Each ADC combines a target-binding monoclonal antibody with a chemical linker and a highly potent cytotoxic payload. Consequently, the safety profile is influenced not only by each component but also by how those components interact. Beyond antigen binding and tissue expression, other factors, such as linker stability, drug-to-antibody ratio, payload release kinetics, nonspecific cellular uptake, catabolism, and systemic free payload exposure, also play a role. And attributing toxicity findings to the specific component causing it is essential to manage those safety risks. Toxicology programmes must evaluate the ADC as an integrated whole while also distinguishing the contributions of the antibody, linker, and payload to dose-limiting toxicity and translational risk.
The importance of toxicity attribution
Because ADC-related toxicities may arise from different parts of the molecule or from how it is processed in vivo, identifying a safety signal is only the first step. Development teams must also determine what is driving that signal. This makes toxicity attribution a central part of ADC toxicology strategy. Effective mitigation depends on whether the risk is linked to the payload, linker, antibody, target biology, or nonspecific uptake.
- Payload-driven toxicity is often reflected in class-specific findings such as myelosuppression, neuropathy, hepatotoxicity, ocular toxicity, or gastrointestinal injury. Mitigation may involve dose reduction, DAR optimisation, schedule modification, or enhanced clinical monitoring.
- Linker-driven toxicity may occur when instability in circulation leads to premature payload release and increased systemic exposure. In these cases, mitigation may require improved linker stability, alternative linker chemistry, or other design changes that reduce free payload exposure.
- Antibody-driven toxicity may result from on-target/off-tumour binding in normal tissues. These risks may require organ-specific surveillance, refined patient monitoring, or, in some cases, reassessment of the target itself.
- Uptake-mediated toxicity can arise through nonspecific internalisation, Fc receptor interactions, or tissue catabolism, and may require optimisation of ADC design features that influence tissue distribution, cellular uptake, or systemic exposure.
By keeping these mechanisms in mind, drug developers are better equipped to identify the toxicology endpoints that will drive a successful and efficient development strategy.
The toxicology endpoints that drive strategic development decisions
For ADCs, the most useful toxicology endpoints are not simply the ones that generate the most data. They should help development teams make better decisions about risk, dose, schedule, monitoring, and candidate viability. A strong ADC toxicology strategy should therefore organise endpoints around three practical questions across a few key phases of testing: What is the ADC doing to the body? Where is the risk emerging? And how should those findings guide clinical development?
- Establish the baseline safety profile The first layer of assessment should define overall tolerability and identify early signs of systemic or organ-specific toxicity. General health observations, body weight, food consumption, clinical pathology, safety pharmacology, and histopathology provide the foundation for understanding whether the ADC is broadly tolerated, which tissues are affected, and whether findings are severe, reversible, or potentially dose-limiting. These endpoints are especially important for identifying target organs of toxicity and determining whether a candidate has a viable safety margin.
- Focus on modality-relevant risk domains ADC programmes should then prioritise organ systems most likely to be affected by the payload class, target distribution, linker stability, and known platform liabilities. Haematologic, hepatic, neurologic, ocular, pulmonary, gastrointestinal, and skin-related endpoints are often especially important. However, not every domain carries equal weight in every programme. A microtubule inhibitor payload may place greater emphasis on neuropathy and myelosuppression, while a topoisomerase inhibitor payload may require heightened attention to gastrointestinal and pulmonary findings. Target expression in normal tissue may also elevate the importance of organ-specific monitoring.
- Connect toxicity to exposure and mechanism Toxicokinetic endpoints are central because they help explain why toxicity is occurring. Measuring total antibody, intact ADC, conjugated payload, and free payload can indicate whether findings are associated with the intact construct, premature payload release, cumulative exposure, or systemic free payload. This exposure-toxicity relationship is essential for interpreting DLTs, assessing linker performance, and distinguishing payload-related toxicity from target-mediated or uptake-mediated effects.
- Translate findings into clinical strategy Finally, dose-defining and recovery endpoints bring the safety dataset together. Measures such as MTD, DLTs, NOAEL, HNSTD, and recovery after dosing help determine the first-in-human starting dose, escalation approach, dosing interval, safety factors, and monitoring plan. In this sense, endpoints are not isolated outputs; they are the evidence base for deciding whether, how, and under what conditions an ADC should advance.
Defining dose, schedule, and first-in-human readiness
Once the core toxicology dataset is established, the next question is how those findings translate into a safe and informative clinical starting point for first-in-human trials. For ADCs, this requires more than identifying a tolerated dose in vivo. Development teams must determine where toxicity begins, which findings prevent further escalation, how exposure relates to those findings, and whether the proposed dosing interval allows sufficient recovery between doses.
Dose-defining measures such as NOAEL, HNSTD, STD10, MTD, and DLTs provide the foundation for this assessment, but they are most useful when interpreted alongside toxicokinetic data. Metrics such as Cmax, AUC, cumulative exposure, time above a relevant toxicity threshold, and circulating free payload can help distinguish peak-driven toxicity from cumulative effects, premature payload release, or schedule-related toxicity.
These data collectively define the ADC’s safe operating range. They inform the first-in-human starting dose, escalation increments, dosing interval, safety factors, and early clinical monitoring strategy, helping teams avoid a trial design that is either too conservative to generate meaningful data or too aggressive to protect patient safety.
The toxicities most likely to shape clinical monitoring and programme risk
Not all safety findings carry the same development implications. After toxicities are identified, the strategic question becomes whether they can be managed within a repeat-dosing clinical setting. Some toxicities are expected, dose-dependent, and monitorable. Others may emerge late, worsen over time, recover slowly, or create disproportionate risk because they are difficult to detect before becoming clinically significant.
ADC programmes should therefore evaluate toxicities not only by organ system, but by their impact. Haematologic and gastrointestinal toxicities may be common and often manageable with monitoring, dose holds, or schedule adjustments. Hepatic injury, pulmonary toxicity, ocular findings, neuropathy, and skin toxicity may require closer scrutiny when they are severe, cumulative, poorly reversible, or linked to target expression in normal tissues or known payload liabilities.
The key issue is not simply whether a toxicity occurs, but whether it is predictable, monitorable, reversible, and compatible with continued dosing. Those characteristics determine how much risk can be managed clinically and whether a safety signal becomes a programme-limiting concern.
A final word
For ADC developers, toxicology endpoints are more than required study outputs. They are strategic tools for understanding mechanism, defining a safe operating range, and anticipating clinical risk. Used well, they can help promising candidates move into the clinic with greater confidence and fewer avoidable setbacks. More broadly, this kind of disciplined safety strategy is essential to translating complex drug modalities into therapies that are not only scientifically innovative, but clinically meaningful for patients.