Continuous manufacturing techniques are not new. Back in the days of the Industrial Revolution they allowed products such as paper to be made more quickly and in larger quantities. Although the concept made inroads into the oil refining and chemical manufacturing space more than 100 years ago, it only started to be genuinely explored in the pharmaceutical sector around the turn of this century. Starting off in academic research laboratories, the idea was gradually assimilated by R&D groups in some of the larger pharmaceutical companies, and momentum is continually building.

The first example of a drug product made via continuous manufacturing was in 2015, when the U.S. FDA gave the go-ahead to Vertex’s cystic fibrosis combination drug, Orkambi (lumacaftor/ivacaftor). Janssen was not far behind, with the approval of HIV drug Prezista (darunavir) in 2016. Both companies have made significant investments in production lines dedicated to the manufacture of these final drug products.

The use of continuous technology to manufacture active pharmaceutical ingredients (APIs) is following close behind. While several organizations have made investments to drive innovation in this field, the industry leader is arguably Eli Lilly and Company, which established an industry-first continuous manufacturing line at its site in Kinsale, Ireland, where a three-step continuous process to synthesize the investigational cancer drug prexasertib is being used to manufacture clinical trials supplies under GMP conditions.

Advantages over batch
One major advantage of continuous over batch manufacture, is that it can expand the available processing space, both in terms of capacity and capabilities. Effectively shrinking the size of the necessary equipment not only reduces the footprint of the manufacturing space required, but also makes key process parameters such as temperature and pH easier to control within the chemical process. This enhanced control can manifest itself in a reduction in impurities, and therefore, an improvement in product quality, by minimizing the likelihood of side reactions, which are common in large batch reactors.

Safety is obviously another big driver. As the reactions are carried out on a far smaller scale, energetic chemistries and toxic compounds can be handled with far less risk, in contrast to the much greater hazards that are inherent in a large volume batch reactor. Furthermore, a potentially unstable intermediate can be made in situ and then used immediately in the next step, with the quick associated timescale minimizing the safety and quality impacts of degradation, and may eliminate them entirely. Similarly, some final products are unstable to the reaction conditions required to make them, and continuous flow reactors are generally designed to minimize residence time under such harsh conditions.

Additionally, commercial quantities for newly developed drugs may be less than that of their counterparts. For example, some drugs are highly potent, or have a limited target market with relatively few patients. Small commercial batches are extremely expensive to carry out in large-scale equipment because of high infrastructure costs, as well as verification and cleaning procedures still being required. Developing a continuous process makes the use of dedicated, or even disposable equipment for small volumes more feasible, which could have an impact throughout the entire supply chain.

Further exploring the commercial benefits of continuous manufacturing, there are notable examples of companies making significant investments in facilities and equipment in the late stages of the development process, only for the project to fail in Phase III. Investment in continuous processing represents a lower risk approach, although costs may be slightly higher for process development, in the long term they could well be significantly lower. The same equipment that was employed for process development and pilot-scale manufacture could potentially be used for commercial manufacture, minimizing, or even avoiding, the additional infrastructure costs, while streamlining process validation efforts in terms of both time and expense.

Another big impact on potential timelines involves the scaling up process. Having to scale up multiple times, from laboratory scale to pilot scale for clinical trials, and then up to the final commercial scale can be very time-consuming and costly. With a continuous process, the scale up becomes much easier, as the scaling factor is much smaller, and may not exist at all: the equipment that is used for the process development becomes the commercial equipment. A large portion of the development cycle can thus be significantly reduced, if not completely eliminated, which is a strong argument that advocates for the inclusion of continuous flow in the early portions of a product’s development.

Additional process development gains can rise from the freedom to explore a wider process window from the onset. Standard commercial batch equipment can have certain limitations in terms of temperature or pressure capabilities, and as a result, the process and product development is forced to operate within those same limitations. Performing the process in flow may remove such constraints at the outset, and more extreme conditions such as elevated temperatures and pressures can be explored to see if they might accelerate the kinetics of the process, while reducing the risk of stability issues. This broader window of potential parameters can be a powerful tool in the development phase, and can allow a process to be optimized from a chemistry perspective. If the reaction can be made more selective and make fewer side products as a result of fine adjustments to the conditions like this, it will make the whole process more efficient and will have a markedly favorable impact on process intensity by reducing, or eliminating, additional purification steps that will be required later.

Applicable technologies
While in theory any synthetic process could be carried out in continuous mode, in practice, the chemical reaction must be relatively quick. If the reaction time cannot be reduced below a few hours, it is unlikely to be a good candidate for moving away from batch manufacturing. That aside, most current technologies have an analogue in the continuous world, and flow also opens up some types of chemistry that are not scalable using current batch processes. Photochemistry is an obvious reaction type that is generally not compatible with large batch reactors, but in continuous flow there is a real opportunity to use this chemistry in a production setting.

Continuous processing really comes into its own with those types of reaction that are difficult or dangerous to do in batch reactors. The most obvious example would be energetic chemistry, as it is unwise to use large quantities of explosives in a batch reactor. Other good candidates include steps involving highly toxic reagents and intermediates, where the need to handle or isolate them can be removed, and reactions where very high temperature or pressure are necessary. If a process has a very tight operating window where any deviations from ideal conditions will negatively impact quality, this will also be a good candidate, due to the enhanced control capabilities.

Multistep processes
Another key advantage of continuous flow is the ability to connect multiple reactor modules together so that several steps can be carried out without isolating intermediates. In some cases, it is as simple as connecting one module to the next in series, where the output of the first step is the direct feed for the next, without the yield impact of isolating the product. This is particularly advantageous if this intermediate is toxic or genotoxic, by reducing the safety precautions that will be required. It is also ideal if the intermediate is unstable, again having a positive impact on yield and quality. A further cost advantage is that the equipment and time that would otherwise be required for isolation is unnecessary.

In some instances, direct connection of multiple steps may not be feasible due to logistical challenges and changes in product flow, but these issues can be solved by introducing a surge tank to catch the product from one step, where it can accumulate for a short time before being fed into the next step, facilitating the overall multistep process. This also gives some contingency in the event of a process upset or mechanical failure, while also providing a natural point in the process for lot identification and traditional off-line sampling and analysis.

It is also possible to engineer out the potential for cross-contamination between products, which can be a particular issue in a multipurpose facility that is not dedicated to a single product. Although the cost of small-scale reactors and equipment is certainly not negligible, it is significantly lower than the investment required for batch-scale machinery. But if an ingredient or process is highly destructive to the equipment or poses a high risk of contamination, it might be feasible to consider the equipment effectively as disposable. This is clearly not the case with a large, multipurpose batch facility, which must be thoroughly cleaned before running a different process.

Some chemistries are notoriously known as ‘black magic’ chemistry, sometimes failing for reasons that are unknown. Committing high value materials into batch production, where the possibility of failure is high, is not an appealing proposition. However, continuous flow can help ameliorate this problem. Real-time monitoring of product quality from a flow reactor allows sub-standard materials to be diverted to a hold tank for recycling or disposal. Once the parameters are tuned and the reaction is working well, a clean, good quality product can be collected. 

And the future?
Now that continuous processes are starting to make inroads into the pharmaceutical market, it seems likely that many more APIs will be made in this way in the future. More and more pharmaceutical companies are turning to continuous processing for developmental drugs in their pipeline. As more and more compounds reach commercial launch, companies will publish their positive continuous flow results, perhaps inspiring others to follow in their footsteps. 

Regulatory agencies are becoming more supportive of flow chemistry as well, and discussions between industry leaders and regulators are ongoing, providing valuable insight about key parameters such as what constitutes a ‘batch’, and what testing and validation data will be required.

These developments in continuous flow certainly do not spell the end of the road for batch chemistry, which will continue to play an important role in the manufacture of APIs. Instead, batch and continuous manufacturing will co-exist, as there will always be some processes that are better suited to batch mode.  

But as we continue to develop new drugs, especially those for niche, targeted indications where only small quantities of API will ever be required, and complex chemistries where continuous manufacturing makes the process simpler, quicker or safer, flow chemistry will gain in importance. Production in continuous flow is an important tool in our ever-growing synthetic toolbox, and is certain to become part of the wider arsenal for development and production of APIs in coming years.