Antibody-drug conjugates (ADCs) are one of oncology’s most productive modalities because they promise something every drug developer wants: potent tumor destruction with greater selectivity than conventional cytotoxic therapy. As our understanding of ADCs matures, one reality is becoming harder to ignore. In ADC development, the antibody may direct the molecule, but the payload often defines the real therapeutic index. An ADC’s payload directly drives cytotoxicity and shapes both efficacy and safety through its mechanism of action, potency, and physicochemical properties.
That distinction matters a great deal for scientific accuracy. It also changes how teams should think about portfolio risk, candidate selection, and IND-enabling strategy. A payload is more than the active ingredient attached to an antibody; it is often the primary source of recurring toxicities, a major determinant of free-drug exposure risk, and an important driver of the nonclinical evidence package needed to support first-in-human (FIH) testing. When those realities are understood early, drug developers and sponsors have options. When they are discovered late, the cost is usually measured in timeline delays, test re-designs, and last-minute scrambling.
Precision Delivery, Constrained by Payload Biology
The commercial success of ADCs has focused attention on antigen selection, linker stability, and conjugation technology. Those variables remain essential, but the source review provides a useful corrective: payload evolution has always been an exercise in balancing lethality against safety. First-generation payloads like methotrexate and doxorubicin entered the market with early ADC development during the 1980s to early 2000s. These payloads were hindered by relatively weak cytotoxicity, hydrophilicity, poor membrane permeability, and significant systemic toxicity. Second-generation payloads (2011-2013), led by auristatins and maytansinoids, dramatically improved potency and enabled many of the marketed products that established ADCs as a viable platform. Third-generation payloads (post-2019), including topoisomerase I inhibitors and PBD dimers, pushed activity even further, often with strong bystander effects and reduced recognition by efflux pumps.
From a development perspective, that progression is not just historical context. It shows why payload selection has become a strategic decision rather than a chemistry detail. As payloads become more potent and biologically mobile, the key question is no longer whether the ADC can kill tumor cells. The primary consideration is whether the payload release profile, membrane permeability, and tissue exposure can support a development path that remains coherent under nonclinical and clinical scrutiny.
Why Payload Class Predicts Liability
Marketed ADC payloads fall mainly into two broad categories: microtubule inhibitors and DNA-damaging agents. Each class carries recognizable liabilities that should shape both preclinical strategy and clinical planning.
Microtubule inhibitors such as MMAE, MMAF, DM1, and DM4 are among the most familiar payloads in the field. Auristatins interfere with microtubule dynamics, induce G2/M arrest, and trigger apoptosis; MMAE combines strong membrane permeability with a bystander effect, which may support efficacy while also contributing to peripheral neuropathy and hematologic toxicity. Maytansinoids show overlapping toxicities, but ocular events, including corneal issues, are more prominent in some programs. These are not incidental observations. They reflect the way payload mechanism intersects with normal tissue vulnerability, and they provide an early clue as to where dose-limiting toxicity may emerge.
DNA-damaging payloads, on the other hand, create a different set of opportunities and challenges. Topoisomerase I inhibitors such as DXd and SN-38 have expanded ADC potential in solid tumors, but the review ties topoisomerase I payloads to interstitial lung disease risk, particularly when strong membrane permeability enables diffusion into lung tissue and initiates epithelial injury and inflammatory cascades. Calicheamicin derivatives and PBD dimers deliver extreme potency through DNA double-strand damage or irreversible cross-linking, but the same features that make them attractive for resistant disease can also narrow the safety margin if release or exposure is not tightly controlled.
For sponsors, the business implication is clear: payload class offers an early forecast of liability. That forecast should inform not only molecule design, but also target product profile assumptions, biomarker strategy, safety monitoring plans, and the scope of IND-enabling work.
Recurring Toxicities Are Development Signals
A common mistake in ADC development is to treat toxicity data as a late-stage challenge instead of an early-stage signal. Drug developers must strive toward a more integrated view by linking recurrent adverse events to payload behavior. Hematologic toxicity, especially neutropenia, is the most common finding and is mechanistically linked to injury in actively dividing bone marrow cells. Peripheral neuropathy is repeatedly associated with microtubule inhibitor payloads because disruption of axonal transport is a direct extension of mechanism. Ocular toxicity has been associated with MMAF- and DM4-containing ADCs, potentially through corneal epithelial exposure by systemic distribution or tear-associated routes. Interstitial lung disease remains one of the most strategically important liabilities in topoisomerase I inhibitor programs because it can alter both benefit-risk perception and development trajectory.
For teams preparing for IND, these are important clinical management issues. They are also translational signals that help answer tougher questions. Specifically:
- Is the payload too diffusible for the intended biology?
- Is the linker-release mechanism creating an unexpected free small-molecule burden?
- Are the observed toxicities consistent with known mechanism, or do they point to a more complicated exposure story?
These questions matter because they influence everything from dose escalation planning to the credibility of the safety narrative presented to regulators and investors.
Novel Payloads Expand Opportunity & Burden of Proof
The review is especially valuable in its discussion of emerging payload classes because today’s scientific promise most often intersects with tomorrow’s development risk. The field is moving beyond established cytotoxins toward immunomodulators, RNA inhibitors, Bcl-xL inhibitors, proteasome inhibitors, NAMPT inhibitors, and other novel mechanisms. These payloads may improve selectivity, address resistance, or extend activity to quiescent tumor cells, but their safety liabilities are often less well characterized than those of established payloads such as MMAE, DM1, or DXd.
That novelty changes the nonclinical strategy. For well-characterized payloads, toxicity evaluation can often be accomplished through studies on the intact ADC. For novel payloads or payloads with unclear toxicity profiles, however, a separate toxicological evaluation of the payload itself is needed in relevant species, either as a standalone study or within a dedicated arm of the ADC toxicology study. It also highlights separate hERG assessment and genotoxicity testing as typical requirements for novel payload components. Reproductive toxicity, photosafety, metabolite assessment, and carcinogenicity considerations should be addressed case by case, guided by mechanism, indication, and existing evidence under frameworks such as ICH S6, ICH S9, and related guidance.
This is where experienced infrastructure and cross-functional planning start to matter disproportionately. Novel payloads can create differentiated assets, but they also compress the margin for error. If payload-specific liabilities are identified early, developers can still refine linker strategy, exposure margins, species selection, and monitoring plans. If they surface late, the same innovation that once looked like differentiation can become the source of avoidable program risk.
A Final Word
ADC innovation is rapidly moving toward more complex payload biology, stronger bystander effects, and multi-modal mechanisms designed to improve efficacy and overcome resistance. The opportunity is real, but so is the challenge. Payload choice does more than determine how an ADC kills a tumor cell. It shapes recurring toxicities, influences the adequacy of the nonclinical package, and can determine how resilient the development strategy will be under FIH expectations.
The bottom line is that sponsors cannot be conservative when it comes to innovation. Instead, they should be more deliberate about where risk sits and how early it must be characterized. In ADCs, late toxicology surprises affect timelines, budgets, clinical momentum, and strategic options. Early payload-specific insight, by contrast, creates room to optimize, de-risk, and move forward with greater confidence. At the end of the day, IND readiness means a complete package and a development story that is scientifically credible, operationally executable, and built to withstand scrutiny.
