ADC payload and linker supply: where capacity runs short

Where do ADC payloads and linkers come from?

Not from the antibody supply chain. They come from small-molecule chemistry, usually at a different company.

ADC manufacturing splits into four steps that need different equipment and different expertise: making the antibody (cell culture and purification (in Japanese)), synthesizing the payload and linker, attaching them to the antibody (conjugation), and formulation and filling. Because the skills differ so much, the steps normally end up at several companies. Each split adds a technology transfer (in Japanese), with specifications to translate and material to ship in between.

There are three main ways to get a payload-linker. For research and early studies, you buy a commercial product, such as the ADC payload-linker products in our directory. For clinical and commercial supply, a contractor synthesizes it for you under GMP. The third way is to make it in-house on your own platform. Daiichi Sankyo reuses its DXd payload across several products, changing the target each time (see the DS1025 news item).

Some suppliers are trying to cut the number of handoffs. Indena promotes an integrated European platform that combines fermentation, plant-derived raw material sourcing, processing under high containment and GMP scale-up. In Japan, FUJIFILM Toyama Chemical plans to start in 2027 an integrated service that takes antibody, conjugation and drug product all the way through domestically (Fujifilm and Taiho partnership). Putting everything with one company reduces handoffs, but it also concentrates your supply in that one company (make or buy (in Japanese)).

Sourcing payloads and linkers means finding a chemical synthesis and containment contractor, not buying an antibody.

Why is a drug too strong to use alone made in dedicated suites?

Because containment decides the equipment, and the equipment decides the economics.

A payload is not a medicine by itself. It becomes a usable treatment only after an antibody carries it to the target. The flip side is that the plant handles a highly toxic compound in its raw form. The amount an operator may safely inhale is many orders of magnitude smaller than for a typical API. That allowable amount sets the containment class (OEB) (in Japanese), and the class sets the equipment specification. Lonza says that an expansion it completed in 2022 at Visp, Switzerland, can handle compounds with an exposure limit of 1 ng/m³, which is one billionth of a gram per cubic meter of air.

In practice that means isolators or closed transfer systems, dedicated air handling and exhaust, a dedicated waste route, and separate entry and gowning for people. Once a suite is built to this level, there is little reason to make a low-toxicity product in the same room. Whether a room can be shared is decided from each API's health-based exposure limit (HBEL), but for highly pharmacologically active compounds the answer in practice tends to be a dedicated suite.

Volumes are also small. Most of an ADC molecule's weight is the antibody, and the drug attached to it is a small part of the total. Heavy equipment and a tiny output is a combination that depreciation never covers.

The compound is not impossible to make; a facility this strong rarely earns its cost back, and that is what holds back new entrants and expansions.

Why does capacity grow by adding rooms, not by scaling up?

Because every payload and linker has its own synthesis route, so there is no standard process to scale.

Antibodies share a template: express in CHO cells, capture on Protein A. Payloads and linkers have none. The starting materials, the reaction conditions and the purification are reworked for each product. Making the reactor bigger to get more output works poorly, so an expansion takes the form of lining up manufacturing suites of the same size side by side.

On June 30, 2026, Lonza announced that it would add commercial-scale capacity for HPAPIs and ADC payload-linkers inside an existing GMP building at Visp. The layout lets it add suites as demand requires, and it expects the expansion to run in 2028.

Adding suites in parallel does not cut the investment per suite. Each suite needs its own qualification and its own cleaning validation. Costs stack up with every addition, so it is hard to commit while demand is unclear. When you ask a CDMO about capacity, ask how many products it can run at the same time rather than the total reactor volume. That question tends to get an answer closer to reality.

Payload capacity is measured by how many dedicated suites exist, not by the size of the reactor.

Where does the quality line sit: starting materials and impurities?

Earlier than most buyers expect, and it depends on what your analytical methods can measure.

Impurities formed during payload-linker synthesis can themselves be highly pharmacologically active. The more toxic an impurity, the lower its control limit has to be. Then the detection limit of the analytical method becomes the first constraint. Before you fix a specification, check that a method exists that can measure the value you propose. That is the correct order.

The second issue is the starting material, the point where GMP control begins. ICH Q11 asks you to explain how it was chosen, based on knowledge of the actual commercial process. If the synthetic route is long and an intermediate along the way is already highly pharmacologically active, the discussion moves to controlling the earlier steps under GMP, and the number of contractors that can do it shrinks. A starting material accepted in one manufacturer's process is not necessarily accepted in another's, so changing contractors can mean explaining your choice again.

Payload-linkers are small molecules, but they end up inside a biologic. Which regulatory framework applies to them is something to settle with the regulator early.

The steps after conjugation also check that no free drug is left. A linker that detaches easily raises the free drug. How a drug that detaches in the body behaves is covered in ADC PK/ADME (in Japanese), and the toxicity a payload can cause in ADC nonclinical safety and payload toxicity (in Japanese).

The more toxic the impurity, the lower its limit tends to be, so a method that can measure that low is the precondition for the specification.

What does a 2028 start date mean for a contract you sign today?

Not that your product can be made there in 2028. A start date for an expansion is not a date for your batch.

It takes years for a facility to be built, qualified and for its processes to be validated. That forces developers to reserve capacity before it is clear whether the clinical results will hold. Long-term capacity contracts with CDMOs exist to bridge this time gap. If the program fails, what you paid is lost. If you do not reserve, you cannot make the product. If you wait for the clinical results, the slots that are open at that point are years away.

Decide on reserving slots according to the equipment queue, not only the prospects of your molecule.

What to ask your supplier or CDMO before you commit

Whether you work in purchasing, QA or manufacturing, put these questions to the supplier when payloads and linkers come up.

1. Will you buy the drug-linker (the payload with the linker already attached) or the two separately? Who guarantees the quality of the intermediate before conjugation?
2. Is the stated capacity a number of rooms or a reactor volume? How many products can run at the same time?
3. What is the containment class (OEB), and is the suite dedicated or shared? If it is shared, what is the HBEL basis?
4. Where is the starting material set, and who supplies the material before it? If you change contractors, are you prepared to explain your choice again?
5. When a company says it can handle "ADC end to end", which of antibody, payload-linker, conjugation and filling does it do itself, and which does it subcontract?
6. Is there an analytical method that can measure the impurity limit? What is its detection limit?
7. Does the capacity-reservation contract state the cancellation conditions and how unused slots are treated?
8. Does the contract require advance notice of changes to raw materials, manufacturing site and method?

If you are in sales, ask first which payload family your customer's ADC uses. It tells you the containment level and the range of testing that will be needed.

Commercial payloads and linkers can be compared in the directory pages for ADC payloads, ADC linkers and conjugation solvents. For how the conjugation method changes the picture, see site-specific conjugation (in Japanese).