From bench to plant: What really changes during scale-up

Word count: Approximately 2,000 words
Estimated reading time: 10 minutes

Dr Gillian Gill discusses the pitfalls and challenges often seen during scale-up, and how working with an experienced CDMO partner can mitigate risks through a phase-appropriate manufacturing approach.

Process chemistry is just medicinal chemistry in a larger flask, right?

Wrong.

While it is true that medicinal chemistry and process chemistry are both vital parts of drug development, each represents a different phase of a drug’s development journey and requires a distinct approach to planning and executing chemistry to deliver high-quality material. These fundamental differences relate directly to the phase’s primary objectives.

During the medicinal chemistry – or drug discovery – phase, numerous screening reactions are undertaken with the primary intention of finding promising candidates through hit-to-lead and subsequently, lead optimisation approaches.  The ultimate goal is to synthesise a range of compounds to test against an assay or target, and the synthetic route to the required compound may be designed to introduce late-stage diversity rather than optimised for high yield, atom efficiency and reproducibility.

Process chemistry requires an alternative approach because the target compound is a known entity: the lead candidate has been identified, and now a robust, large-scale process to manufacture the clinical candidate active pharmaceutical ingredient (API) is needed. The process chemists will likely use the original synthetic route devised by the medicinal chemists as a starting point, but will work to identify improvements that deliver an efficient, reproducible, economic, safe and environmentally sustainable approach to large-scale production.

While the term “process chemistry” implies a one-size-fits-all approach, a standardised approach is rarely the best option. Within process development, the exact phase for which material is required – preclinical, Phase I, Phase II, Phase III or Commercial – directly impacts how a synthetic strategy is developed. This means process objectives can (and will!) change as a molecule progresses through the development pipeline. Good practice dictates a phase-appropriate approach. At a minimum, from the preclinical phase and up to Phase I, this involves ensuring existing processes are safe to scale up while improving efficiency and throughput. The route will also be challenged with the caveat that challenging stages may be retained on a cost/risk basis if the alternatives risk delaying hitting Phase 1 milestones.

Generating reliable yields and delivering materials of high chemical purity to satisfy an evolving specification is a must. Processes and the intended route are further fine-tuned as the candidate progresses through development, driving continuous improvement to the manufacturing procedures. As a candidate nears Phase III or Commercial Manufacture, techniques such as Design of Experiments (DOE) and Failure Mode and Effects Analysis (FMEA) are conducted to both improve process knowledge and understanding and to identify critical parameters to further derisk these larger-scale – and more costly – phases.

The breadth and depth of experience at Onyx Scientific mean our approach to process development and scale-up is well tested and proven.  We have significant experience improving processes to reduce costs, enhance efficiencies, and shorten timelines. Furthermore, because Onyx routinely works across the development timeline, we know where pinch points and challenges can arise and actively work to future-proof our clients’ synthetic approaches to avoid costly redevelopments between process phases.

This blog post will consider the key aspects to consider when scaling a process from bench to kilo scale.

Solvents matter: Volume, identity and substrate stability

During a medicinal chemistry campaign, reactions often run in high volumes of solvent relative to the substrate. However, when it comes to developing a scalable process, solvent volume is a crucial aspect for consideration – not only in terms of process throughput and efficiency, but also in relation to environmental impact. That said, reducing solvent volume is not necessarily simple. Aspects such as the solubility of starting materials, reagents and products must be considered. In addition, viscosity becomes increasingly important because it affects stirring and mixing, and close monitoring of both mass and heat transfer is vital to ensure safety and minimise by-product formation. However, solvent volume is not the only variable, and consideration of the solvent itself is also important. In many jurisdictions worldwide, controls on solvent use are increasing – especially when it comes to chlorinated or fluorinated solvents, as well as those that are carcinogenic and teratogenic. In all cases, the International Council for Harmonisation (ICH) rating should be considered, favouring Class III.

A common misconception is that dissolving a substrate in solvent is without risk. This is not true. An assiduous CDMO will ensure stress tests are performed to confirm chemical and thermal stability under the proposed reaction conditions, especially for extended hold times and longer vessel residency required by larger-scale processing.

Materials matter: Reagents, starting materials and catalysts

All chemical reactions include one or more starting materials and reagents and, for large-scale processing, a diligent CDMO will consider cost, grade or quality, and supply reliability. What goes into a reaction directly impacts what comes out: quality and supplier control are key.

Ensuring starting materials, reagents and intermediates are qualified as suitable for use vs a specification is critical. The impact of reagent stoichiometry will also be investigated to avoid unnecessary excesses. Controlling stoichiometry can also limit side and/or by-product formation, directly cutting costs by reducing purification requirements and limiting waste, with significant environmental benefits.

Catalyst screening is important to identify the best catalyst for a reaction and to optimise catalyst loading. Cost and availability of the required catalyst will also be considered, as larger-scale reactions can require large quantities of catalyst even at low loading. If a catalyst is expensive, is not commercially available, or has a long lead time, asset delivery to a client can be directly impacted. In many cases, the question is often asked: “Is this the best catalyst for the reaction? Does another cheaper, or more commercially viable catalyst exist that gives comparable reaction outcomes?” While in some cases the answer may be ‘no’, in others, a balance may be struck between reaction outcome, cost and time.

Phases matter: Stirring and mass transfer

While monophasic reactions can often be optimised through use of concentration effects, thermal control or reagent selection, biphasic reactions can be much more capricious – especially on scale-up. In a biphasic reaction, efficient mixing between phases is essential, with agitator type, rate and vessel dimensions playing an important role. Gas–liquid systems, such as those used in hydrogenation, can also be significantly impacted by mass transfer. In this case, using a traditional hydrogen blanket over the substrate solution may lead to sluggish reaction and poor conversion, especially at larger scales where the interfacial surface area is smaller. However, a specially designed impeller that introduces gas directly into the reaction mixture through sparging can increase reaction rate, improving efficiency and throughput. For processes involving slurry maturation of a solid suspension, stirring rate is generally listed as a critical parameter for ensuring processes work as effectively at large scale as they do at small scale. Particle size and agglomeration are other quality attributes that significantly impact this type of process and require control through specification setting.

Modelling a (perceived) problem: Safety

The larger the reaction, the larger the risk. In this context, consideration of safety is paramount, and checks and balances – especially in relation to safe limits of scale-up before a safety reassessment is required – are crucial. For example, exotherms must be controlled, and mitigations must be in place if things do go wrong: venting and head-space for off-gassing must be appropriately designed, and the reaction’s thermal profile across reagent addition, reaction, and quench must be understood. Once established, prepare and test the maximum operating temperature and detailed quench procedures, and assess the risk posed by the use or potential generation of energetic materials.

Overall, process chemists must be aware of the fundamental principles of chemical engineering when scaling-up reactions, to ensure that optimum parameters are applied and that any safety issues – such as thermal runaway or pressure build-up – are avoided. Using smaller-scale, process-applicable jacketed vessels with overhead stirrers can help mimic large-scale manufacturing reactors. Deployment alongside modelling software is a valuable exercise before running a reaction in the plant, allowing critical process parameters to be assessed, balanced and adjusted as the process moves from development to manufacture. Through in silico techniques, vessel set-up, size and mixing parameters can be optimised (particularly important for biphasic reactions), endo- or exothermic events anticipated, and heating, cooling and stirring rates all modelled. Furthermore, solvent switches and crystallisation parameters can also be modelled, which can help de-risk these processes on scale-up.

Closing out: Reaction monitoring, and product isolation/purification

Finally, monitoring reaction progress is essential, providing insight into reaction extent and any side-product or by-product formation. TLC, HPLC and 1H NMR are all analytical techniques utilised in process laboratories to monitor reactions. Regardless of the method used, the overall goal is to optimise the reaction profile to ensure consistent batch purity (including all individual impurities) while maximising reactor throughput.

Product isolation and purification involve different considerations in drug discovery versus process chemistry. At full process scale, product isolation requires more thought. Generally, extractions and washes using high volumes and successive manipulations are not favoured, and the purpose of each wash must be fully understood and justified based on the reaction components being removed. Where possible, isolation by precipitation or crystallisation is preferable, as this can minimise the time and processing required to remove solvent and can itself act as a useful purification step. Once the target material has been isolated, attention can then turn towards further purification.

In contrast to drug discovery, where column chromatography is often the first (and only) approach to product purification, process-scale column chromatography is generally avoided because of the large solvent volumes and the time required. Several other techniques, including (re)crystallisation, salt formation, acid/base washes, scavenger treatments, slurry maturation, and in-line filtration, are viable alternatives that can often be deployed quickly and cheaply with excellent results. However, regardless of the technique used, assessing chemical purity and tracking impurities to ensure they fall within final product specifications is vital and often requires a range of analytical techniques.

The Onyx advantage

Working with an experienced CDMO partner, such as Onyx, provides significant gains. With experience supporting numerous clients with their small-molecule manufacturing projects, the Onyx team is well-versed in solving process-development challenges and optimising syntheses. As a result, we are likely to have encountered — and resolved — many of the scale-up and process issues commonly seen during development.  Furthermore, our ability to manufacture materials from preclinical to commercial large-scale means we are always looking ahead, ensuring compound manufacture is phase-appropriate and meets the client’s needs.

Planning your next scale-up?

Whether you are preparing for your first kilo-scale batch or refining a process for later-stage manufacture, considering scalability early can help reduce risk and avoid costly redevelopment. Contact the Onyx Scientific team to discuss how we can support your process development and scale-up requirements.

Frequently Asked Questions

When scaling from med. chem. to process chem., what aspects do I need to consider?

There are many aspects to consider, including: reaction concentration and solvents used; reactants, reagents and catalysts; safety – for example, thermal or pressure build-up, or energetic materials; and purity of the final API. Scale-up is not necessarily a matter of simply adopting the medicinal chemistry route and often requires re-evaluating the route and adapting to deliver scalability and robustness. Talk to us today to find out more.

Are all process chemistry phases the same?

No! As a molecule progresses from preclinical through to commercial, the synthetic route and approach taken necessarily change. Good practice is to use a phase-appropriate approach, fine-tuning operations as a candidate progresses through development. Talk to us today to find out more.

Why should I work with Onyx?

As a small molecule CDMO with broad experience across large-scale manufacturing from Phase I through to commercial, Onyx is well placed to support our clients across complex manufacturing projects. Furthermore, because we work through to commercial manufacture, we set ourselves apart from competitors and ensure processes are future-proofed, avoiding costly redesigns and route redevelopment during later stages of large-scale manufacture. Talk to us today to find out more.