EXECUTIVE INTERVIEW - Lonza: Driving Innovation & Sustainability: CDMO Trends & the Future of ADCs

As the pharmaceutical industry evolves, Contract Development and Manufacturing Organizations (CDMOs) like Lonza are at the forefront of addressing complex challenges and driving innovation. From advancing antibody-drug conjugates (ADCs) to embedding sustainability into manufacturing practices, Lonza continues to set benchmarks in delivering solutions that meet the needs of today’s healthcare landscape.

Drug Development & Delivery recently interviewed Lonza’s Sebastian Stenderup, Executive Director, Head of Commercial EMEA, about some of the trends CDMOs are currently facing, from the growth of ADCs to the importance of sustainability, and how his company is responding to them.

Q: What do you think will be next in the world of ADCs?

A: One aspect that is coming into sharp focus right now is the payload. About half of the antibody-drug conjugates (ADCs) that are already on the market use one or other of two payloads – and most target microtubules or inhibit DNA-acting enzymes. What might be next?

A good deal of research effort is being put into the next generation of payload-linker combinations. For example, there is significant interest in looking again at older and less potent payloads for traditional ADCs.

Two of the ADCs that have gained significant market penetration target topoisomerase 1, both of which use derivatives of camptothecin. What about payloads targeting topoisomerase 2 instead? While earlier attempts to create them proved unsuccessful, there may be potential, and multiple other payloads with diverse mechanisms of action are also being investigated. It is going to be interesting to see what the next big thing will be.

Away from the payload itself, modified linker technologies could improve the therapeutic index by targeting different lysosomal enzyme triggers that avoid payload release in healthy tissues. There should be plenty of mileage in developing modified payload-linker combinations, including finding ways to improve their release from the ADC once they reach the target, or even incorporating two different payloads in the same ADC. And if the drug-antibody ratio can be better controlled, and the conjugation made at a more stable site on the antibody, this could also lead to more effective ADCs.

What about ADCs that include bispecific antibodies? These could be designed to allow the antibody to give more specific targeting of tumour cells that co-express two cancer associated antigens that are not present in non-tumour tissues. And, of course, smaller peptides might be able to have a similar targeting effect to antibodies. Peptide-drug conjugates are still in their infancy, but the potential is clear.

Q: What about alternative mechanisms of action?

A: It’s true that the lion’s share of ADC effort has been focussed on traditional cytotoxic payloads, but there are certainly indicators that alternative mechanism of action payloads could have utility in the fields of cancer-immunology and also targeted protein degradation, too. It’s early days, of course, but there is considerable interest in the potential of immunostimulatory ADCs. STING – stimulator of interferon genes – is a great example as it is involved in the tumour-dependent activation of the immune system’s quiescent myeloid cells that can then attack the tumour directly. And while there have been some early setbacks in the clinic, it may be that toll-like receptor agonists (TLRs) could ultimately prove effective as ADC payloads, too.

It could even be that targeted protein degrader molecules such as PROTACs could be conjugated to target-directing antibodies. There is growing preclinical interest here, and as the molecules tend to be fairly complex, experienced CDMOs like Lonza are well placed to contribute in this space. While they are complex molecules, in practice the E3 ligase targeting units they contain are fairly modular, lending themselves to libraries.

Q: I guess the complexity poses additional challenges for GMP manufacture?

A: Definitely. One is obvious – the different volumes that are required depending on the application and the nature of the components. Some ADCs are being developed to treat solid tumours with high prevalence, where the quantities needed will be much greater than for those designed to treat rare haematological cancers. Volumes are also affected by the potency of the payload – there can be a 2-log difference between the potency of different payloads, and the more potent it is, the less will be required. The drug-antibody ratio can also significantly impact volumes, too.

In order to cover the entire space for ADC manufacture, a range of GMP capacities will be required, from a few tens of grams for niche products in the early stages of clinical development through to hundreds of kilograms for the commercial manufacture of ADCs targeting more common tumour types where the market is, naturally, larger.

Containment requirements pose a significant challenge because of the high potency of the molecules. A CDMO looking to work across the range of ADC types and from preclinical through to commercial will have to have capacities at multiple scales. With the requirement for OEB5 containment, this represents a significant investment, and one that Lonza has made to ensure we can support customer projects across the board.

Q: What about the physical nature of the molecules? Can you give me examples of challenges there?

A: A good example is the crystallinity of payload-linkers, which is often problematic. Many contain solubilising or hydrophobic-masking chains such as polyethylene glycol (PEG) or polysarcosine. These impact the crystallinity of the payload-linker combination, making them more complicated to purify. We often find that large-scale preparative HPLC is required, followed by lyophilisation. And, of course, both of these will need to be carried out under OEB5 containment. This has a significant effect on productivity, and unless those time-consuming purification and isolation steps are taken offline, they can seriously impede the overall productivity of an asset by reducing its availability for other projects.

Q: Aggressive development timelines are becoming increasingly common, so presumably this is also the case for ADCs?

A: Absolutely. The industry is increasingly promoting DNA to IND times of 13 to 15 months, and this includes Lonza. In reality, for ADCs to achieve these timelines would require a GMP payload-linker to be fully developed in eight to 10 months. And, of course, there’s far more to a payload-linker structure than simply the small molecule API component. The growing complexity of ADC payload-linkers means they typically include a solubilising linker, a protease trigger, a self-immolation motif and a conjugation handle in addition to the small molecule payload itself.

This timescale can only be achieved if a well-defined supply chain is in place for all the building blocks. Otherwise, developing, scaling up and manufacturing the payload-linker in fewer than 10 months would be extremely unlikely. An additional layer of complexity comes with the decision on which building blocks will need to be sourced as GMP starting materials.

Q: How can high-throughput experiments help?

A: High-throughput experimentation is an essential tool in finding the optimal process as quickly as possible, whether the project involves ADC payload-linkers or more traditional small molecule drugs. By carrying out test reactions on a small scale and in parallel, we can get pointers towards the conditions that might be successful far more rapidly. For early phase projects, these are invaluable – whether working on route selection, or screening catalysts, ligands and solvents.

A robotic system can carry out far more experiments in parallel than a human chemist could ever hope to, and with a broader range of conditions that could be used in large-scale manufacture, such as high temperature or pressure. It allows us to narrow down the options for solvents, reagents and reaction conditions far more quickly by screening all the many different options in parallel. The human chemists can then get to work refining the conditions, and carrying out scale-up experiments that are less amenable to automation.

Importantly, our high-throughput robot in Visp is supported by a dedicated team and includes UPLC analytics capabilities. Analytics can frequently become a substantial bottleneck in high-throughput operations, and we are keen to prevent this from slowing our development projects down.

Q: This leads us nicely on to route scouting. Can you tell me more about Lonza’s new AI-enabled route scouting service?

A: Sure. Route scouting is all about designing the optimal way to make a target molecule. It builds on the decades-old technique of retrosynthetic analysis, where a chemist looks at a complex organic molecule and works out how it might be put together from smaller pieces, a bit like a jigsaw. They work backwards until all the pieces needed to assemble the molecule are commercially available.

The potential drugs entering the development pipeline are getting increasingly complicated, and it’s not unusual for a synthetic route coming from medchem to have 20 or more steps. This is slow, resource-intensive, and costly. Artificial intelligence and machine learning are proving a real godsend in trying to find shorter routes that are synthetically feasible, suggesting multiple options for process chemists to triage.

Here at Lonza, we are working with a computer-assisted retrosynthesis tool, which includes a huge database of reactions. We have combined this with our own supply chain database to give commercially relevant insights into the relative costs of different routes, and the real-world availability of the intermediates and reagents. It helps customers to identify and prioritise the best route for manufacturing their molecule right from the outset, building on advice provided by computer tools and the decades of experience of process chemistry that has been built up by our subject matter experts.

Q: With sustainability now such an important topic for all aspects of business, can you give me an example of how Lonza is putting sustainability at the heart of its operations?

A: Sustainability is a broad concept, of course, but looking at the environmental aspects, we are committed to cutting our emissions, reducing our water and energy use, and improving our material intensity. Obviously, this covers a huge number of activities, but one concrete example from our manufacturing operations involves wastewater.

As we must not discharge any traces of API in our wastewater, traditionally it has been incinerated to ensure no API ends up in the local water treatment facility. But this uses a huge amount of energy, so our team in Visp looked at alternative ways that would remove all traces of API while also having a much smaller carbon footprint.

This involves looking closely at the precise contents of the waste stream so we can decide how best to clean it up. It might involve adsorption or extraction, stripping and distillation, thermal hydrolysis, or even membrane technology. And we may need to use more than one technique – perhaps first evaporating any traces of low-boiling solvents, and then adsorbing any API that remains. We’ve also been working with external partners on alternative techniques such as an advanced oxidation process and enzymatic decomposition.

We’ve also been looking to reduce our solvent use. Organic solvents are used in significant volumes in pharmaceutical manufacturing, and represent a substantial waste stream. But the volumes also mean that if they can be recycled rather than disposed of, it will significantly improve the overall sustainability of the process. We have invested in multiple technologies to clean up and recycle our waste solvents, such as distillation, membrane technologies and rectification systems in our sites in Visp and Nansha. There is also a dedicated solvent recovery plant in Visp and, of course, we re-use high-grade recycled solvents within our processes, or sell them on for reuse elsewhere.

These two activities are already making a huge impact on the sustainability of Lonza’s manufacturing activities. In 2023 in Visp alone, we treated more than 2,000 tons of wastewater to avoid incineration, and regenerated more than 10,000 tons of solvent. This is a concrete example of how Lonza has made significant investments to make our manufacturing activities more sustainable.