Filing a Multistage Continuous Process for API

Revolution

In recent decades, the pharmaceutical industry has witnessed a revolution in the regulatory requirements to obtain approval for the manufacturing processes of new medicines. The reliance on end-product testing, was replaced with a paradigm where the control of product quality was designed into the process. Accordingly, when the process variables are held within the designed and studied ranges, end-product testing is now just one of many controls that provide assurance in the supply of consistently high-quality medicines.

Quality by Design

This change in thinking was called Quality by Design or QbD; a concept already well established in the engineering industry and is based on the evidence that most quality issues arise from a poorly designed process. QbD has proven effective in many industries and has been widely used in the automotive industry.

A key part of QbD was the concept of a Design Space; this was defined by the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (shortened to ICH) to be “the multidimensional combination and interaction of input variables (e.g., material attributes) and process parameters that have been demonstrated to provide assurance of quality”; see guidance within ICH Q8.1

Continuous Processing and QbD

Whilst the pharmaceutical industry worked out what this change meant for them and their ways of working, it stimulated an interest in the use of new manufacturing and Process Analytical Technologies (PAT’s). Signalling an increasing desire to move to a new quality paradigm, where quality was controlled in real time. In my view, continuous processing is the ideal technology/process partner to embody QbD.

Running a process continuously means that each unit operation is run in a different piece of equipment, which can be specifically designed for that purpose. Each continuous unit operation runs a unique set of parameters required for that part of the process, at that time. In the example of a chemical reaction, the continuous reactor affords each molecule with a stable environment with consistent mixing, heating and reaction time. This condition is called steady state, where all the process parameters are in a controlled and stable condition; this has been further defined as state of control in FDA guidance on continuous processing.2

Continuous Versus Batch

In comparison to continuous technology, the batch vessel hasn’t evolved very much for the last 100+ years; it is still a stirred pot that is heated or cooled. However, the batch vessel has many advantages because it is a general-purpose tool, the Jack of all trades. The batch vessel can be a tank for dissolving starting materials, for performing reactive chemistry, for quenching the reaction, for performing liquid/ liquid phase separations, for performing distillations and acting as a vessel for crystallization. However, a general-purpose vessel, cannot be the master of any of these unit operations.

A batch process runs in a dynamic manner; the temperature profiles generally change throughout the process and mixing characteristics change as the fill levels change or stirrer speeds are turned up or down. Operating a large (>10,000 litre volume) batch vessel in this manner does not provide the opportunity to control parameters to the same capability as a small continuous reactor (<1 litre volume) running the reaction conditions in a stable regime.

Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!

The Hurdles to Implementation

Given the many advantages for continuous processing, why has the uptake been so slow by our industry? The reasons are complex; apart from the cost of investing in a new plant when the industry has an over capacity of depreciating batch assets, the appetite to use new and perceivably unproven manufacturing technology for a potential block buster molecule, is not high.

In my experience, focusing solely on a successful technical proof of concept is not enough to precipitate the adoption of new technology. The road to implementation requires leadership and a clear vision to align your organization in readiness for change. Underestimating the effort required to influence the broader community in your organization, will only prolong the adoption and allow pockets of change resistance to become embedded.

Regulatory Risks?

The pharmaceutical industry is surprisingly risk adverse and cautious to adopt new ideas. A lot of time has been spent procrastinating over what the risks might be, rather than accelerating the adoption of continuous processing. In December 2019, the multistage continuous process we previously filed, was approved by two major regulatory agencies. Rest assured that whilst there is still plenty of progress to be made, there are no additional regulatory hurdles to implementation.

The current QbD framework and ICH guidance remain applicable when designing and implementing continuous manufacturing processes to API’s. The current and forthcoming guidance3 for continuous processing of API’s, drug products and biologicals, are helping to further define expectations and to encourage more submissions.

My own journey to file a multistage process to API started 20 years ago. As the leader of the API flow chemistry team, for a major pharmaceutical company, I was involved in the design and commissioning of a multi-purpose pilot plant, created to be a demonstration rig for API continuous processes that my team and I developed. During an 8-year period, we developed and successfully demonstrated numerous multistage continuous processes to every New Chemical Entity or NCE we chose to work on. However, due to the high levels of attrition typified by working in the early stage portfolio, none of these NCE’s became products and therefore didn’t make it onto a manufacturing site.

Whilst continuous processing can enable the scale up of difficult chemistry to expedite a clinical milestone, the real opportunities are seated in manufacturing. To that end, I would say that continuous processing represents a true paradigm shift for manufacturing, whereas it is merely another R&D tool; more on this later!

In order to make the benefits tangible in a manufacturing environment, I chose to redevelop an established product, from batch to continuous; ongoing external regulatory engagement proved to be a vital part of this program.

External Regulatory Engagement

In September 2013, I presented the project to a group of senior leaders at the FDA headquarters as part of a short meeting on new technologies. By 2015, the process was nearly fully developed, and we engaged with the European Medical Agency (EMA) PAT team in September and with the FDA’s Emerging Technologies Team in December. During this period the factory rig - see Figure 1 - was installed at the manufacturing site and commissioning had started. The team and I embarked upon a couple of landmark meetings; we invited the FDA’s ETT to the manufacturing site for a 5-day pre-operational visit during August 2016. We held a similar 4-day meeting with representatives from the PMDA in March 2017. The EMA PAT team were invited to the manufacturing site for a similar preoperational visit, but due to budgetary restrictions, they were sadly unable to accept the invitation.

Looking back, it is clear to me now that one of the major benefits of the pre-file engagement was to drive internal alignment; a date booked with a regulatory agency does tend to focus the attention of any organization, precipitating important discussions ahead of engaging externally. Many other useful discussions took place at these regulatory meetings and much of the pre-file engagement themes were echoed in the post-file questions.

One example of this was the use of PAT and the role it would play within the control strategy. It is worth taking a moment to set the scene to fully appreciate the root of these questions. This was the first multistage continuous process within the manufacturing network for the company. For this reason, it was a natural decision to include the same suite of PAT tools we had employed in R&D, to the manufacturing rig. Thus, helping to facilitate a smooth technical transfer by enabling a comparison of manufacturing performance against earlier R&D data.

However, herein laid the problem; the data from the traditional PAT equipment (online HPLC and conductivity probes) was only collected for comparative purposes, or for information only. Our control strategy and design space was predicated on a multivariate combination of input material attributes (specifications), the control of the process parameters for the continuous chemistry and the batch isolation of the API, in combination with all the usual cGMP controls, under the umbrella of the Quality Management System. The control strategy was not influenced in anyway by the data being collected by the PAT instruments. Therefore, the role of PAT was challenged throughout the regulatory process, because there is no such thing as for information only in the manufacturing environment.

In this example, I would propose that the Coriolis mass flow meters and numerous temperature probes could be considered to be true PAT tools; they are fundamentally measuring a process parameter and providing an input to the control system, precipitating an action to maintain the parameter within the design space.

The remaining regulatory questions common to both pre- and postfile were largely operational in nature, seeking assurance that the usual checks and balances were in place. By way of example, the process utilizes two Divert to Waste (DTW) valves that are triggered if a process parameter value exceeds the design space. The valves do not fire automatically after a process disturbance; instead a timer is started based upon the Residence Time Distribution (RTD) of the process train.

When the timer expires, the control system opens the DTW valve just before the material made in the process disturbance arrives. Thus, the quality of the material in the collection pot is maintained and the amount of material diverted to waste is minimized. To further explain the background to these operations, additional data was requested to explain the characterization of the reactor train and how the RTD timings were used to trigger the Divert to Waste valves.

The dossier(s) were submitted to FDA and EDQM toward the end of 2018, approvals were obtained from both agencies in December 2019. The FDA approved the process without a Pre-Approval Inspection; one can only surmise why, but the level and quality of the pre-file engagement may be a contributing factor.

Next Steps

Now that a multistage API process has been filed and approved within the pharmaceutical industry, it is important to reflect again on why the pace of change is so slow. I have been privileged to collaborate with many of the major players within this industry.

From my perspective, most companies are looking to their R&D departments to take the lead and to choose where, when and how to deploy the technologies within a new route to a new asset. Whilst it is appropriate for R&D to select the best route and evaluate suitable technologies, this approach is not going to deliver any significant pace of change within the API industry, due to the attrition of R&D projects; it is a one-brick-at-a-time approach.

Accepting that the continuous plant is cheaper to build, occupies a smaller footprint and uses less energy than a comparative batch facility, what are the benefits for the patient? Most major companies have looked at the cost of poor quality within their organizations. Moving from batch to continuous provides a step change in the precision of process control and the ability to manage quality in real time. Making this a reality requires a strategy driven by manufacturing rather than R&D. Perhaps, a realistic first step for API manufacturers, would be to target the redevelopment of troublesome impurity forming stages of existing assets, into continuous processes.

About the Author

Dr. Malcolm B. Berry has worked in the pharmaceutical industry for 28 years and spent 26 years of his career in API process development. In the last two decades, his focus has been the pursuit of continuous processing for API’s. He has recently started his own company MB Chemistry Consulting Ltd, working as an independent consultant in the design and implementation of continuous processes. Additional information is available at https://mbchemistryconsulting.co.uk/ and https://ngtbiopharmaconsultants.com/

All the views expressed in this article are of the author and they are a summation of his experiences working across the industry and are not representative of a former employer. For author correspondence, please write to [email protected]

References

1. International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH), Pharmaceutical Development Q8, August 2009
2. FDA draft guidance: - Quality Consideration for Continuous Manufacturing Guidance for Industry, February 2019.
3. ICH Q13. According to ICH, the proposed new quality guideline will:

• • Harmonise Continuous Manufacturing related definitions
  • Articulate key scientific approaches for Continuous Manufacturing
  • Harmonise regulatory concepts and expectations for Continuous Manufacturing across the regions