For all the progress in antibody-drug conjugates, two biochemical challenges have constrained the field for much of its history. Both are, at their core, chemistry problems — and both are being addressed through increasingly sophisticated linker-payload engineering. Mersana Therapeutics’ emiltatug ledadotin (XMT-1660), a B7-H4-targeted ADC, is a particularly clear illustration of how rational design tackles them head-on.

Two Long-Standing Constraints

Heterogeneity from stochastic conjugation

Traditional conjugation chemistry attaches payloads to the antibody more or less at random, producing a heterogeneous mixture of species with varying drug-to-antibody ratios (DAR). The high-DAR species in that mixture tend to be the most hydrophobic, and hydrophobicity carries a penalty: accelerated plasma clearance, suboptimal pharmacokinetics, and ultimately a narrower therapeutic window. The mixture is only as good as its worst-behaving components.

The bystander-versus-safety trade-off

The second challenge is more subtle. To kill antigen-negative tumour cells in a heterogeneous tumour, a payload needs to diffuse out of its target cell and into neighbours — the bystander effect. That requires high membrane permeability. But the same permeability that enables beneficial bystander killing also enables systemic efflux, where the payload escapes into circulation and causes dose-limiting, off-target toxicity. The property you want locally is the property you fear systemically.

How XMT-1660 Addresses Both

A branched scaffold for uniform DAR

Instead of a conventional linear linker attached at random, XMT-1660 uses a branched scaffold that is site-specifically conjugated to the antibody. This predefined geometry produces a highly uniform DAR — optimised at approximately DAR 6. The resulting structural homogeneity stabilises the pharmacokinetic profile, reduces premature systemic clearance, and improves tumour biodistribution relative to stochastically conjugated ADCs. The mixture problem is solved by not making a mixture in the first place.

A payload that switches its own permeability

The most elegant feature is the payload chemistry. XMT-1660 carries auristatin F-hydroxypropylamide (AF-HPA), a microtubule inhibitor. In its free state, AF-HPA is highly lipophilic and membrane-permeable — exactly what is needed to diffuse into adjacent antigen-negative cells and produce the bystander effect.

But the design does not stop there. After cellular internalisation and cleavage, intracellular enzymes metabolise AF-HPA into auristatin F (AF). AF is highly polar and membrane-impermeable. Because of this, the fully active cytotoxic agent cannot re-enter systemic circulation. The payload, in effect, alters its own permeability state in situ: permeable when it needs to spread locally, impermeable once it has done its job — so off-target toxicity is mitigated while local antineoplastic activity is preserved.

This is a genuinely clever resolution of the bystander-versus-safety dilemma: rather than choosing between the two, the chemistry sequences them.

The Broader Lesson

XMT-1660 illustrates a principle that increasingly defines the field. The therapeutic index of an ADC is not fixed by the choice of antibody and payload alone — it can be engineered. By controlling DAR heterogeneity through linker architecture, and by designing a payload that actively changes its physicochemical behaviour at the right moment, medicinal chemistry is fine-tuning the therapeutic window itself.

It is a strong model for what rationally designed targeted therapeutics can look like: every structural decision serving a specific pharmacological purpose.

How SigutLabs Can Help

This kind of design depends on chemistry expertise at the linker and linker-payload level — exactly where SigutLabs works. We support early-stage ADC innovation through:

• Design and synthesis of custom linker and linker-payload architectures
• Branched and site-specific conjugation scaffolds for controlled DAR
• Payload and prodrug strategies that tune permeability, stability, and release behaviour

If you are developing next-generation ADCs and want a chemistry partner to help engineer your linker-payload architecture, we would be glad to connect.

📩 Get in touch