Building a Continuous Processing Culture

This year, Patheon has officially launched a continuous manufacturing initiative. Its new continuous processing facility in Greenville, NC was completed in September of 2016, and the company’s director of continuous manufacturing, Eric Jayjock, is bringing newer approaches into the company’s toolkit and culture.

Although oral solid dosage (OSD) forms are now the focus, the company’s division in Linz, Austria, offers expertise in continuous API flow reactor chemistry that might be used in the future. Dr. Jayjock discussed plans and thoughts about continuous manufacturing with Pharmaceutical Technology.

Experience developed at Janssen and Rutgers

PharmTech: How did you get into continuous manufacturing?

Jayjock: I was a PhD candidate at Rutgers working for Fernando Muzzio, and my thesis was on a completely different subject involving inhalable drugs. However, during this time, I had developed the ability to combine a chemical engineer’s process knowledge with the building aptitude of a mechanical engineer. 

These two skills came into play when Johnson & Johnson asked Rutgers to help them build a prototype continuous, direct compression line to evaluate the technology for their facilities. This became my post doctorial project and eventually led to Janssen offering me a position. My family and I moved to Puerto Rico in 2013, where I supported the development of a continuous direct compression line for the production of Prezista.  

This time it’s different

PharmTech: A lot is being written about continuous processing, but what makes it different from earlier movements such as process analytical technology (PAT) and pharmaceutical quality by design (QbD), which, initially, were embraced by an elite and only slowly gained traction throughout pharma?

Jayjock: PAT and QbD are highly scientific pursuits, allowing organizations to develop the depth of knowledge that will help them run product more efficiently. At first, companies may have assumed using these approaches would bring some regulatory relief, but that didn’t necessarily play out as some people had anticipated. 
In the end, however, you still make the product the same way with PAT and QbD, so there was no real sea change. From a high-level business perspective, the question became: Where do I save money if I use PAT or QbD? Of course, these methodologies allow you to save money because they will enable you to develop more robust products. Pharma, however, is driven by new products and that’s where most of the industry’s investment goes. There are significant financial benefits to producing products continuously. 

Continuous manufacturing reduces plant footprint and labor, and permits higher capacity utilization. Today, 30% capacity utilization is still typical for many batch pharma facilities, but with continuous pharma, plants will be able to run closer to chemical plant levels, i.e., with 80-90% capacity utilization. That way, every dollar spent on equipment can potentially yield two or three times more bang for the buck.

PharmTech: Is Patheon’s continuous manufacturing line now up and running?

Jayjock: We’re close. We started building the new facility in Greenville, NC in March 2016, and finished it by September 2016, then we had to install equipment and start commissioning various aspects of the system. 
We begin detailed process development in early 2017, adding a supervisory process control system that will integrate all the pieces of equipment into a flexible framework of units. This framework will allow Patheon to use a common suite to manufacture products with different processing needs. So 2017 marks the start of our continuous processing initiative, and our facility is almost ready to begin manufacturing. 

PharmTech: Will Patheon offer continuous for OSD forms, APIs, or both?

Jayjock: Long term, we’re looking at being able to provide continuous manufacturing wherever it makes sense from a business and quality perspective. Right now, my focus is on OSD, but we might want to branch out into other areas at some point in the future.

As you go farther back along the pharmaceutical value chain to the drug substance, it becomes more like a chemical process and less like a traditional pharma stream, and we already have continuous chemistry capability in Linz, Austria, for API manufacturing. Patheon’s group in Linz has scaled up several different reaction pathways using a continuous flow reactor. This approach allows development trials to be run on ,a very small scale, using very low flow rates, and then scaled up via large manifolds. Our colleagues in Austria have been doing this type of work for 10 years now.

PharmTech: From what you’ve seen, are people in pharma skeptical, fearful, or interested in continuous processing?

Jayjock: Branded drug companies with new drug entities are either already on board with continuous or coming on board rapidly, while others are still studying the best way to get into this area. Small- and medium-sized branded pharma companies, for whom CDMOs are the hands of their operations, are very interested.
The work on continuous that has been done by Vertex and Janssen has shown that continuous processing can save a great deal of material because it eliminates the need for scale-up. For high priority, rapid designation product, continuous processing can allow a great deal of drug development time to be saved. In general, there’s been a lot of interest.

PharmTech: How do you educate clients and would-be clients about what will be required?

Jayjock: I’m spending a quarter to 30% of my time doing some educational work, both internally and externally with clients. This type of approach is necessary because continuous manufacturing is so data driven. You can study much more about a process than you could in the batch world, so you can determine what poses the greatest risk to each product. By combining a risk-based program with the data rich nature of continuous manufacturing, we drive a much deeper understanding of our products and processes.

It would be a real disservice to treat this like a traditional, cut and dried batch manufacturing project, as in ‘We’re going to produce X batches of Y size’.  

With continuous processing, we bring our customers’ scientists on board to talk about problems and custom tailor programs to the unique attributes of each product. 

Cultural change required

PharmTech: Are there any fears that clients have?

Jayjock: What I advise companies of all sizes, whether they want to go into continuous manufacturing themselves or are looking to work with a partner, is that it’s not just about equipment, or even technology, but about people. You need to change mindsets throughout the organization and that’s often the most difficult issue.
For example, we tell people they will have to work with their quality teams to show them how continuous -processing is different from batch to help them understand. Typically, people think about the process and technology first. Then, when they become more comfortable with those aspects, they worry about integrating a different approach into the mindset of their organizations.

PharmTech: Are quality departments afraid of continuous, and potential loss of employment?

Jayjock: I would not say afraid, but rather unsure. Continuous manufacturing quality assurance requires a different skill set from that used for batch. Quality for batch manufacturing is set up so that you simply run quality tests, run metrics on the batch and then have a conversation about whether you pass or fail the batch.  
With continuous manufacturing, you see an issue in production that may lead to an automated response by the control system. This is something quality needs to be a part of developing up front during validation. Once you implement control loops to keep your process within specifications. Then a deeper level of quality conversation can take place. 

Let’s say you are producing products which are on spec, but the combination of processing conditions needed to get the product there is different from typical batches. Is this a quality issue? The answer is it is, of course, very product dependent.  So the quality organization needs to be a partner in the development process to determine what is critical for each product. The quality conversations come much up earlier and quality has to be involved from the beginning.

It’s like learning a new language. We see there is hesitancy and concern at some companies, because it is different from what everyone is used to, and they are not sure how to train the new organization.

PharmTech:  How would the way you start a continuous project with a client differ from the approach that you’d take with batch?

Jayjock: The financials are very different. In the batch world, for example, it doesn’t make a big difference whether you have four or eight components in a blend. With continuous manufacturing, it makes a very big difference, because you’ll either need four or eight individual blenders, so there will be a significant engineering difference between the two approaches.

As far as transferring information is concerned, the biggest difference is the traditional batch approach is pretty well scripted. We know from the outset, for instance, that we’re going to make X batches of Y size, and scale to Z, so we can easily predict how much material and time will be needed, and what type of work must be done.
Continuous manufacturing is more iterative, as in ‘We’re going to do this bundle of work, these small scale experiments, to establish whether or not these risk factors are critical to the process. We’re going to study these factors, then we can remove some from the list or prioritize other factors for the next round of study … maybe additional factors will be needed, so there will be a little more work required than we originally anticipated.’  
With continuous processing, there’s a stage gate approach where the scope of work is always changing. The idea is that we’re always looking at risk to the product and making sure we’re putting resources toward the most valuable studies we can run. 

From the business side, it’s a challenge because we can’t quote with the same type of regularity seen with batch manufacturing, so there is no cookie cutter approach. More conversations have to happen in order to determine the scope of work.

PharmTech: How many of your clients are embarking on continuous?

Jayjock: We currently have two projects signed on to the continuous manufacturing train. 

Understanding foundational concepts

PharmTech: How well do most clients understand the foundational concepts, PAT, QbD, and process control, behind continuous manufacturing? 

Jayjock: Companies who are looking for continuous solutions tend to have people working for them, each of whom has one or another piece of the puzzle in place, given the data-driven nature of continuous. For instance, companies may have people with statistics, PAT, or Six Sigma backgrounds.

Our goal is to take something that is, on the surface, highly complicated-a process where there may be 10-15 different parameters we can adjust to keep it in control, with 5-10 material properties that can have an impact on the product-and boil it down to its simplest, easiest to understand, form: ‘These are the two things that matter most to our product, and these are the two things that can control them.’ In pharmaceutical development, we need to have a clear control strategy that outlines risks to the product, and how we can best handle them.

The secret to manufacturing is that only a handful of factors really drive quality. A process engineer’s job is to sort out the bad from the good and determine which ones are most important to the final product’s quality.

Obstacles to real time release testing

PharmTech: What is being done to develop real time release testing (RTRT) for continuous? Janssen has filed for this, but PAT guidance had promised that this would be possible.  What is needed, since dosage forms today are becoming more complex?

Jayjock: Often continuous manufacturing and RTR are put in one category, with the oversimplified idea that you make product continuously and it just goes out the door. There can be certain advantages to doing RTRT for continuous manufacturing in some cases, but not always. There are many more tests we run on batch products for which we have superior surrogates, but you’re not always going to do RTRT on continuous processes.

For more complex formulations, continuous processing gives us a test point. Right now, we have variability in materials, put them into a fixed process and get variable product, so we test it to determine quality.  

With the continuous development cycle, you can vary the process and see the effect on the product in near real time within a few minutes so we can tie that knowledge together. The most valuable thing we’ll have for RTRT is process data. If a product is being produced, with attributes tightly controlled using a combination of process conditions which are statistically comparable from traditional production, you have a lot of data to support the quality of your product. In general, continuous is a great enabler, but not necessary for RTRT.

My personal point of view (not that of my employer) is that we’ve set up this culture of testing for quality, but some tests may have questionable utility, for example, dissolution, which shows little in vivo in vitro correlation. It’s not clear what the test is supposed to be doing, except to measure variation in quality for a particular tablet. Statistically, however, it is not that good at determining variability in quality. Yet we keep using it, despite the fact there are better methods for doing this. The largest barrier for RTRT is that we’re being held to this test, which doesn’t have a lot of statistical meaning.

For some time, the conversation in industry has been: ‘Can we predict what happens in a dissolution test?’ That would make sense if the dissolution test itself were a statistically better predictive measure. We’re spending a lot of energy on dissolution when we could use other measures, such as the physical properties of tablets, that are greater indicators of  product quality.

PharmTech: What are your goals, short and long term?

Jayjock: In the short term (i.e., next six months to year), we’ll be getting the first suite up, and bringing the whole organization along into the technology. We want to bring the company’s culture to the point where we understand and can execute continuous projects as an organization.

Medium term, we want to focus on ensuring suites are multifunctional and batches can be turned over very quickly. The engineering challenges involved are very real, and we have a plan so that we can make continuous better, not just from quality, but from business perspective.

Long term, we want to build a manufacturing framework within which we can roll continuous manufacturing projects out anywhere in the world, and where we have the utilities in place to support them (e.g., to build modular unit operations, ship them and be able to run the process six weeks later after we have them on hand).

PharmTech: Do you have a dedicated staff for continuous?

Jayjock: Currently, three senior level scientists (including myself) on the development team, and we are developing an operations team. 

We staffed the development team this way because the three interaction points are:

• Understanding material properties and variability

• Understanding the process 

• Understanding what’s happening in real time (that’s where the chemometrician comes in) so we can sense changes in the line. 

PharmTech: Which equipment vendors you are working with?

Jayjock: My personal belief is flexibility should be the focus, so which equipment we use is secondary to how we bring it all together (e.g., the piping), which can work for one product but not another. We also need deep process understanding, and a feel for what works well and what doesn’t.

The concept behind our design is we can integrate any equipment into our suite in which the vendor will work with us. For the first wave of work on the new facility we selected equipment we thought would be a good fit for our system: Gericke for continuous feeders; Korsch for compression; Lodige for wet compression and automation, and a Gerteis minipactor for dry compaction. We’re considering adding a new continuous coating system for future expansion. 

PharmTech: How does turnaround time differ for a batch vs. a continuous product?

Jayjock: It is difficult to generalize, but the time required to turn a continuous line around and switch product is not a problem for manufacturing. Where it can be challenging is in drug development, however, when you are running many, short product runs. 

Article Details

Pharmaceutical Technology
Supplement: Outsourcing Resources
Vol. 41
August 2017
Pages: s24-s27

Citation

When referring to this article, please cite it as A. Shanley, "Building a Continuous Processing Culture," Pharmaceutical Technology Outsourcing Resources Supplement (August 2017).