An article by J. J. Douglas and A. D. Campbell et al at Astra Zeneca (AZ) was published in the March 2025 edition of ACS Catalysts [1]. The authors of the paper detailed a fascinating 20-year journey of the implementation of high throughput experimentation (HTE) at multiple sites, and how learnings from one site led to developmental improvements. Their paper is entitled “The Implementation and Impact of Chemical High-Throughput Experimentation at AstraZeneca.” and overviews developments in chemistry, software and hardware (such as CHRONECT XPR), to allow them to realise significant improvements in throughput and approach to HTE.
Developing and launching new drugs onto market is extremely challenging, for example in 2024 only 50 novel drugs were approved by the US food and drug administration (FDA)2. When this is compared to 6923 current active clinical trials (registered by industry at the time of writing on clinicaltrials.gov3), it can be appreciated that new drug approval and deployment rate is very low. Therefore, successfully launching new drugs is risky and extremely expensive. For example, recent study in Frontiers of Drug discovery, estimated that the development pathway of a new medicine, takes around 12-15 years and costs around $2.8 billion from the point of inception to launch4.
One of the most costly and challenging areas of the drug development process is the initial candidate selection and optimisation. The most common approach for this is high throughput screening (HTS), where hundreds of thousands of compounds are screened to identify thousands of potential? hits. Following HTS, lead optimisation is conducted to refine the number down to a handful of candidate molecules for further development intended for the clinical development stage5.
High throughput screening is just one part of high throughput experimentation (HTE). HTE looks to massively increase throughput of all processes employed in the discovery and development of drugs. One key area of this is in parallel chemical synthesis of drug intermediates and final drug candidates. There are two primary focusses in parallel chemical synthesis, the first is optimisation of the synthesis of key drug intermediates, and the second is the synthesis of analogue libraries from late-stage precursors6. As well as enabling massive improvements in throughput, HTE is also conducted at much smaller scales than traditional synthesis - in terms of amounts of reagent and solvent used per experiment. Not only does this aid logistics, pertaining to sample handling and storage, but also working at such small scales, results in a significantly lower environmental impact. For example, traditional round bottomed flasks on heaters are replaced by vials in heated or cooled 96 well array manifolds. These in turn are operated in inert dry atmosphere gloveboxes where robots are able to, work n what would be potentially hazardous conditions for Lab technicians, for both powder and fluid manipulations. Furthermore, automation of these functions also frees up personnel to work on tasks requiring more humanized knowledge and intuition.
The LVE, or Library Validation Experiment, is such an example; during this experiment, in one axis of a 96 well array, the building block chemical space is be evaluated, and the opposing axis, specific variables can be scoped such as catalyst type and or solvent choice (all at mg scales)vi.
The paper by J. J. Douglas and A. D. Campbell et al, overviewed a fascinating 20 year history of the implementation and evolution of HTE at AZ. The article outlines that the team had 5 goals to implement HTE.
1. Deliver reactions of high quality.
2. Twenty catalytic reactions would be screeded per week within 3 years of implementation.
3. Development of a catalyst library.
4. Rather than just achieving reaction ‘hits’, they aimed to understand reactions more comprehensively.
5. Employ principal component analysis to accelerate reaction mechanism and kinetics knowledge.
The team at AZ highlighted specific hurdles they would need to address to successfully employ HTE workflows. One key area of focus was in the in the implementation of automation of solids and corrosive liquids as well as ways to minimise sample evaporation. These were partially and initially solved by use of inert atmosphere gloveboxes, a Minimapper robot for liquid handling employing a Miniblock-XT holding 24 tubes (employing a resealable gasket to prevent evaporation of solvents). Finally, during this initial period, the group used a Flexiweigh robot (Mettler Toledo) for powder dosing stating, “Although in many ways imperfect, the Flexiweigh automated weighing robot was the starting point for the current generation of excellent weighing devices.”
Use of automation at AZ has evolved significantly from these early days to being employed at multiple sites. For example, during 2010 the team at AstraZeneca at Alderley Park UK helped Mettler develop user friendly software for their Quantos Weighing technology. Over the ensuing years, Trajan* and Mettler would then collaborate to develop a next generation powder & liquid dosing and weighing technology called CHRONECT Quantos and further evolved into the modern technology CHRONECT XPR. This combined Trajan’s expertise in robotics using Trajan’s Chronos control software and Mettler’s market leading Quantos/XPR weighing technology. The robot operates within a compact footprint and so enabling the user to handle powder samples in a safe, inert gas environment - critical for HTE workflows.
· Powder dispensing range: 1 mg - several grams.
· Component dosing heads: Up to 32 Mettler Toledo standard dosing heads.
· Suitable powders: Free-flowing, fluffy, granular or electrostatically charged.
· Dispensing time - 1 component: 10 - 60 seconds, depending on compound.
· Target vial formats: Sealed and unsealed vials (2 mL, 10 mL, 20 mL); unsealed 1 mL vials.
One of the key successes in deployment of HTE at AZ was in oncology. In 2022, the team invested $1.8M in capital equipment at both the Boston USA and Cambridge UK R&D oncology departments. For example, CHRONECT XPR systems were installed at both sites, to handle powder dosing.
Furthermore, two different liquid handling systems (one for each site) were also installed. This HTE initiative, led to a significant increase in lab efficiency. For example, at the Boston facility, the average screen size, prior to automated installation (Q1 2023), increased from between ~20-30 per quarter during the previous 4 quarters, through to an impressive ~50-85 per quarter, over the following 6-7 quarters. Even more remarkably, was the ramp up of the number of conditions that could be evaluated (<500 to ~2000 per over the same time).
As well as automation, the authors also reported the importance of collocating HTE specialists with general medical specialists and they stated the following. “We view the colocation of HTE specialists with general medicinal chemists as highly beneficial to the HTE model within Oncology, enabling a co-operative rather than service-led approach adopted by other peer pharma HTE groups.”
According to an article in BioSpace in November 2023, entitled “Biologics Projected to Keep Gaining Ground in Cancer Therapeutics Market”, stated that by 2029 (for just oncology alone), the market size of biologicals is predicted to far outstrip small molecules7. Moreover, a 2025 article published in the journal Molecules, stated that only one in three FDA approved drugs (N=50) in 2024, were small molecules, the others were large molecules included proteins, TIDES (two oligonucleotides and two peptides) and 13 monoclonal antibodies8. It can be appreciated that just from these two facts, discovery of novel biopharmaceuticals is critical to the pharmaceutical industry and so implementation of HTE in this area, is a key tool in allowing for an increase in production of hits and candidate molecules.
As highlighted by J. J. Douglas and A. D. Campbell et al, the importance of HTE in biopharmaceutical drug discovery at AZ, was also a key priority and in 2023, development of a 1000 sq. ft FTE facility was initiated at the Gothenburg site. By building on their prior FTE experience, as well as garnering knowledge from previous laboratory projects, they designed the facility to have 3 compartmentalised HTE workflows (gloveboxes A, B and C). Glovebox A was dedicated to automated processing of solids and employed a CHRONECT XPR automated solid weighing system as well as providing a secure and safe place to store solids including catalysts†. This was particularly important to omit liquid reagents to best preserve the reactivity of catalysts. Glovebox B was dedicated to conducting automated reactions and validation of HTE conditions to gram scales. Glovebox C, was home to standard equipment used in their global HTE teams, where they could, for example, run reaction screening using liquid reagents combining liquid automation as well as options for manual pipetting. This also allowed the team to continue to hone their experience with miniaturisation.
The authors described a case study which showcased the advantages of using the CHRONECT XPR at the AZ HTE labs in Boston. The following benefits were observed.
· A wide range of solids (for example, transition metal complexes, organic starting materials and inorganic additives) was able to be successfully dosed.
· When dosing
o At low masses (sub-mg to low single-mg), < 10% deviation from the target mass was observed.
o At higher masses (>50 mg), < 1% deviation from the target mass was observed.
· A significant reduction in time weighing was observed compared to manually. For example, manually weighing typically took 5-10-minutes per vial. In contrast, a whole experiment took less than half an hour including planning and preparing the CHRONECT XPR instrument.
· For more complicated reactions such as catalytic cross coupling, using XPR powder dosing for 96 well plates scales. This was found to be significantly more efficient and furthermore, eliminated human errors, which were reported to be ‘significant’ when powders are weighed manually at such small scales.
The authors highlighted that much of the hardware for FTE was either now developed, or will likely to be developed, in the not-so-distant future. Nevertheless, the team highlighted a need for significant development in software, to enable full closed loop autonomous chemistry. A feat that they previously demonstrated in within their flow-chemistry labs. Indeed the authors noted that other advances have been made in the field, such as self-optimising batch reactions, but currently these still require a lot of human involvement, in experimentation, analysis and planning. Therefore, they concluded that there is more to be done.
It is well known that in any industry, ‘garbage in’ often means garbage ‘out’. Unless quality is addressed in all stages of a process (hardware and software), then as throughput of a system is increased, quality of the end products often suffers as a consequence. The paper by J. J. Douglas and A. D. Campbell et al, showcased a fascinating 20-year journey that AZ has undergone, to successfully implement HTE in their labs globally. In doing so, they have significantly increased drug candidate throughput but critically at the same time maintained an important level of quality. Using state of the art robots, such as the CHRONECT XPR instrument, is one example of a innovative technology, which has contributed to AZ to successfully implemening HTE in multiple locations.
1 https://pubs.acs.org/doi/10.1021/acscatal.4c07969?ref=PDF
4 https://www.frontiersin.org/journals/drug-discovery/articles/10.3389/fddsv.2023.1201419/full
6 https://pubs.acs.org/doi/full/10.1021/bk-2022-1420.ch001
7 https://www.biospace.com/biologics-forecast-to-continue-gaining-ground-in-cancer-therapeutics-market
8 https://www.mdpi.com/1420-3049/30/3/482