A Continuous Cycle of Success

April 2019 marked the end of Scott Gottlieb’s highly active tenure as FDA Commissioner. Throughout his time in the role, he made it explicitly clear that continuous manufacturing (CM) was a key factor in supporting innovation and “modernizing the pharmaceutical industry” (1). The call has been echoed by many others within the FDA – and elsewhere in the industry too. Pharma companies, however, have been slow to leave batch manufacture behind. Four years ago, only one drug made with CM had been approved and the situation hasn’t improved much since; today, there are around five products approved by the FDA made using CM.

CM undoubtedly requires manufacturers to have a certain level of front-end knowledge in terms of the management of procedures that will result in a successful product launch. Transient solid formation in a batch reactor, for example, is hardly a cause for concern, as the mixture can simply be stirred and eventually the solid should redissolve. But when such a solid forms in a flow process, new challenges arise and manufacturers face delays as they attempt to unfoul their reactors and/or manage the shutdown and cleaning of their systems.

Once these types of teething issues are addressed, however, there are huge benefits to CM. The current batch manufacturing process is costly and lacks flexibility. With a smaller physical footprint, CM enables CMOs and other manufacturers, in principle, to produce final drug products seamlessly in a GMP compliant manner. To date, pharma companies seem more likely to accept CM in regard to tabletting – a low bar for the industry given that many tableting machines already run continuously. I firmly believe that more research effort is needed to improve CM for the development and manufacture of drug substances to de-risk this promising approach and enable broader adoption in the pharmaceutical industry. I’ve heard from many people that although there is strong interest in this area, many projects taking place in companies are running as skunkworks until decision makers can be convinced that it’s worth the investment. It’s also well known that in a highly regulated industry like pharma, it can be difficult and risky to change the manufacturing process for already marketed products. Consequently, it’s clear to me that CM will be more applicable to new chemical entities coming through the pipelines and for generics where the cost of production can be lowered.

It is highly encouraging that the FDA is keen to help companies implement a more continuous form of manufacture into their practice. In February of this year, the agency issued a draft policy that detailed the development and implementation of CM for brand, generic and over-the-counter drugs. It charged its Emerging Technologies Team with the task of aiding early CM adopters with resolving “implementation challenges and navigating the application review process for products made with these modern methods” (1).

Make it work

At Purdue, I am a Professor of Organic Chemistry and my main focus is on drug delivery – designing different kinds of materials that respond to different metabolic conditions to release their drug cargo. It’s very much based on fundamental chemical principles wherein you need to learn about the local environment where you want the drug to be delivered, and then design a material that responds to those conditions. For years, my group and I have focused on the use of microfluidic synthesis of particles for drug delivery. In many cases, the novel synthetic materials we made were mixed via microfluidic devices. As part of this work, we learned to precisely control the size of nanoparticles and designed continuous processes for preparing broadly different classes of nanoparticle assemblies. Since then, we’ve pivoted to also apply continuous synthesis approaches to the telescoped preparation of small molecule APIs.  

As part of a Defense Advanced Research Projects Agency (DARPA) funded project called Make-It (a program designed to help automate small molecule discovery and synthesis), my colleagues and I developed a cost-effective, time-efficient CM method for the manufacture of lomustine (2). Lomustine is a chemotherapeutic used in the treatment of glioma, melanoma, lung cancer and lymphoma.

For Make-It, DARPA envisioned a system that could use artificial intelligence to plan and optimize synthesis routes, and interconnected fluidic modules for continuous synthesis – including algorithms for automation and process control, and in-line characterization and purification (3). Ideally, an individual would be able to request a molecular structure and the system would find out the best method to synthesize the molecule and scale up to tonnes per year routes, and conduct the formulation and tableting. It was very ambitious. Because of the production rates that were called for – and the fact that the desired target small molecules listed by DARPA were multiple step reactions – it was clear that a continuous process would be needed to achieve the desired throughput.

We were inspired to focus on the production of lomustine after reading an article written by Henry Friedman, a neurooncologist at Duke University, in The Cancer Letter. The article explained that the price of the generic therapeutic had increased by 700 percent between 2012 and 2017, despite the fact that each year in the US, 33,000 new cases of lymphoma are diagnosed. Today, the price increase has reached 1400 percent. Consequently, some patients are being priced out of their medication and others are left with the difficult decision of reducing their doses.

DARPA actually provided a list of targets for the Make-It program – and lomustine was not on that list. But in time they enabled our group to expand the focus. At the same time, they also de-emphasized the formulation and tableting aspect of the project. It struck us that if we were able to produce lomustine through CM, we would be able to illustrate the power of a continuous process in catering to the needs of patients marginalized by the ever-growing prices of drugs. We’ve actually looked at four compounds: diphenhydramine, Atropine, diazepam and lomustine – the equipment was the same for all four, underscoring just how flexible CM technology can be.

I assigned one of the students in my lab, Zinia Jaman, to develop a continuous synthesis for lomustine in February 2018. By August 2018, she had created a CM process that produced a highly pure product (higher than the commercial substance) with an overall yield of 63 percent and a residence time (the average length of time to produce the entity end-to-end) of nine minutes. The lomustine was prepared using two separate telescoped flow reactors in a linear sequence. This microfluidic approach was selected for its ability to radically reduce the cost of manufacture and make use of inexpensive starting materials. Most profoundly, the production rate was approximately 110 mg per hour using two coupled reactors the size of a microscope slide. I think this really highlights the power of CM. We can produce a much-needed medicine in a very cost-effective way – and Zinia was able to design this process relatively quickly!

And why stop at lomustine? The approach can also be applied to the production of other medicines, giving us and others within the industry the opportunity to more thoroughly address patients across disease indications.

Based on our work, we have created a company called Continuity Pharma that aims to develop GMP continuous processes for different APIs. Once our milestones are met, I hope we’ll be able to help with one of the biggest issues affecting drug development – and the more we can showcase the advantages of CM, the more we expect the industry as a whole to seize the opportunities that it can so readily provide.

We are also helping colleagues in academia who have interesting drug leads that they need to scale up. It’s not difficult for medicinal chemistry researchers to make enough drug product for mouse studies, but as soon as they progress to follow-up studies on promising leads in larger animals, they need more compound – and many labs aren’t set up for this. We’ve started working with colleagues to scale up their synthesis via a continuous approach to support the progression of these leads through the developmental pipeline.

Counting the benefits

I think the research from my group demonstrates how nimble academia can be when embracing CM and we have begun thinking more broadly about the real-world use of this technology. We can’t say right now what the price will be for CM-derived lomustine because, although we’ve costed the raw materials (around $5 per gram), we haven’t taken into account other elements such as hardware, maintenance, and personnel costs. And the real benefit of CM lies not necessarily in cost, but in production agility. Consider the fact that the output of conventional batch processes leads to the generation of huge quantities of drug intermediates and products that then sit in warehouses waiting to be shipped to the next unit or formulator, leading to an unnecessarily expensive use of capital. For some drugs, this can be a huge problem since not all APIs can be stored for long periods of time without degrading. Being able to produce a drug product just before it is shipped would ultimately enable us and others to improve drug quality and deploy resources more effectively.

The industry is still finding its feet when it comes to CM, but I believe that we will inevitably become more reliant on continuous processes. There is increasing pressure on drug prices and a greater desire for local facilities. Additionally, I think both industry and consumers are becoming increasingly aware of the environmental costs of drug manufacture, particularly with respect to energy consumption and waste generation. Continuous processing allows for high throughput manufacturing in a very small footprint. There is also a lot of discussion about the “ballroom concept” for manufacturing – that is, a large manufacturing area with no fixed equipment, allowing for increased flexibility and plug-and-play functionality. It also reduces the costs of construction associated with multi-story facilities that harbor large reactors.

CM will not replace batch manufacturing practices. Batch allows for high levels of quality control and the ability to modify it to suit company needs allow it to remain an attractive option for manufacturers. But, as CM gains more traction it should be able to work harmoniously alongside more conventional processing approaches.