Continuous Processing—Finally

Though well established in other industries, continuous processing is only now, and only slowly, making its way into the collective pharmaceutical-manufacturing mindset, which is still holding on to traditional, yet reliable, processing methods. There is nothing wrong with most batch processes, but they aren't always the most efficient approach. And, now more than ever, efficiency counts, and continuous processing has the potential to redefine efficiency and quality.

Even as the industry remains rooted in batch processing, there is growing interest in continuous processing. "It certainly has changed," says Paul Mort, principal engineer at Procter & Gamble (Cincinnati, OH) "Ten years ago, my impression was that the pharmaceutical industry held little practical interest in continuous processing. More recently, however, there is a lot more curiosity, especially regarding the implications of continuous processing for process scale-up and control."

The case for continuous processing

The advantages of continuous processing are hard to miss. At the top of this list is the reduced machinery size to achieve throughput equivalent to that of much larger batch units, especially blenders.

"The equipment required for 100-kg/h pharmaceutical processing for a directly compressible tablet is just a couple meters long, including feeding devices, the continuous blender, and an analytical device. The entire system is small enough to be placed over a tablet press for direct discharge," says Tom Chirkot, PhD, PE, laboratory manager at Patterson-Kelley (East Stroudsburg, PA, a division of Harsco Corporation).

Smaller-sized equipment leads to less concern about process scale-up. A batch process can undergo three to five steps of scale-up from the beginning of development through to commercial scale.

"For continuous processing, I could conceivably go through one step of scale-up. I would have equipment to meet my supply needs to Phase 2 and even Phase 3. Then when I need to go the next scale-up, I would do that one scale-up and then use that second equipment and time as my lever to make as much as I need, based on supply-chain demand," explains Mayur Lodaya, PhD, associate research fellow at Pfizer (Ann Arbor, MI).

The key is that time, not capacity, determines the "batch." Continuous processing is inherently advantageous for a product that is in very high demand because the supply chain dictates how many hours the system will operate, increasing or decreasing according to demand.

Continuous processing is not always better than batch systems. There are certainly cases in which a batch process may be the better choice such as when high throughput is not required. "If you have something that you need to run virtually every day that month, or for campaigns that run for a week at a time, continuous processing is ideal," says Chirkot.

"If you require a very uniform residence-time distribution, you may be better off with a batch process. For example, in a coating operation, you want all the particles or tablets to have the same thickness of the coating," says Mort. "The choice of batch or continuous operations may depend on the production scale, control requirements, and availability of raw materials for early-stage development. For example, instead of scaling up from small to large batch unit operations, it may be more efficient to use small-scale batch operations for initial development work and then scale up to a moderate-scale continuous system for production."

Challenges

Despite these benefits, convincing manufacturers to implement continuous processing principles will not be easy.

"Smaller firms may not be as aware [of continuous manufacturing] as some of the larger pharmaceutical companies," says Moheb Nasr, PhD, director of the Office of New Drug Quality Assessment, CDER. "We have made a serious attempt at the agency through our spectrum of initiatives to facilitate and encourage people to utilize such an approach. I don't know if there is an additional need of the agency to provide further clarity of the developed guidances to encourage people to embark on this approach."

Technical hurdles. Coating as a unit operation already has been successfully implemented as a continuous process for neutraceutical products, but there are still some glitches that need to be worked out before it can be used for ethical pharmaceuticals. In particular, the product diversion during start-up and shut-down of the process needs to be defined clearly.

The key difference between a batch process and a continuous process is that in the batch mode, the process is in a dynamic state from the beginning to the end. Depending on the process, the endpoint is predetermined so that when that point is reached, the process is stopped, and the unit operation is done. A continuous process must undergo an initial start-up phase before reaching a steady state. Only after this point can material be collected for further processing and finishing into the final product. The same scenario occurs during the shut-down phase. Determining what to do with the material during these phases and engineering a process that will take into account these scenarios have become primary objectives for those seeking continuous processing systems.

How to handle material that does not meet specifications is another issue. "In commercial processes such as foods, detergents, and fertilizers, manufacturers typically recycle the material back into a continuous process. In fact, some applications require significant recycle loops for process stability. In pharmaceutical manufacturing, companies will have to decide whether to scrap this material or reintroduce it in a controlled way," says Mort.

Recycle loops in pharmaceutical production may complicate abilities to trace the product back to its raw materials.

"A batch of material can be traced as it goes through the process steps," says Mort. "But with a recycle stream going in a continuous granulation process, some materials are recycled one or more times through the process. The recycle history can vary depending on whether the recycle is classified as fine or oversize. Oversize material is sent back through a grinder where it is subject to additional work; or if the material is too fine or dusty it may be collected in a dust-handling system, and now you are faced with the issue of handling fine, cohesive material. In either case, the recycle is separated from its initial cohort of raw materials. Depending on the breadth of the distribution, the recycle streams may constitute a significant amount of material that is spread out though time in the process."

Moreover, in a well-mixed batch system, all of the material in the batch has the same residence time. In a continuous system, the material has a residence-time distribution and the effect of the process-time distribution on the product properties must be controlled. "The need for dose precision is high in pharmaceutical production, so you have to take extra care and go to lengths to make sure your in-process distributions are narrow and in control," says Mort.

Nontechnical challenges. The initial cost of equipment, process development, and quality systems is definately a hindrance.

"Because there is not a requirement to use continuous manufacturing or process analytical technology, the business case needs to be developed to justify the investment of these technologies, which I would argue if they are used correctly and appropriately it may result in considerable savings and enhance the efficiency of pharmaceutical manufacturing," says Nasr.

Perceived regulatory issues and current financial constraints of some manufacturers also are delaying the introduction of newer technologies.

Some believe it may take competitive pressure from generics to make pharma pay greater attention to manufacturing efficiencies. "In my opinion, it is not conducive for them to make things continuous because there is no economic incentive," says John Kossik, technical manager, Steadfast Equipment (Mill Creek, WA). "The processes for pharmaceuticals and biopharmaceuticals are in many cases not any more complicated than those used in the chemical-process industry. The essential problem with getting continuous processing into prescription drugs is that the laws of supply and demand don't work anymore, especially in high-value drugs, so there is no motivation to be efficient."

For a process that already has been developed, there is capital required to change it to a continuous mode. "If a company has a high-value drug they already have approval for, there is no reason for them to switch over to continuous processing until the economic conditions change. When drugs go off patent even in large pharma companies, they make process changes because they have to compete with generic-drug manufacturers, especially now that insurance companies are 'encouraging' people to buy generics," says Kossik

Progress

Still, there are signs of progress, most notably within FDA's continuing CMC pilot program. Begun in the fall of 2005, the program provides pharmaceutical manufacturers with an opportunity to submit applications that demonstrate modern approaches of pharmaceutical development or quality by design (QbD). As part of this pilot, the agency works with inspection teams, reviewers, and compliance officers from CDER to facilitate the review and the implementation of QbD principles for companies that submit an application. So far, the program has received more than 12 submissions and has accepted 11 into the pilot and as of last month approved at least three applications. These applications demonstrate the concepts of PAT such as immediate-release and some elements of continuous processing.

Although implementation issues still need to be addressed, Nasr says he has witnessed "tremendous progress on the industry's part as far as its appreciation and some very serious attempts to implement some of the new approaches in manufacturing and quality systems."

Advancements in clinical production also are notable. "We are just starting to work with innovator pharmaceutical researchers on the clinical manufacturing side to introduce continuous processing into their pilot-lab sequences," says Chirkot.

Lodaya admits the industry is still in the "evolutionary phase" of continuous processing for the pharmaceutical industry, but that it is an area under active development and analysis. "To my knowledge, all major pharmaceutical companies have programs working in this area. Over the next five to seven years, I think all the major pharmaceutical companies, those that have the technical power and the ability to put people into these projects, will have their own way of using continuous processing," he says.

Applications

Companies already have taken steps in applying continuous or semicontinuous strategies to some unit operations.

Blending. Patterson-Kelley developed its "Zig-Zag" continuous blender, an adaptation of the common V-type mixer, a rotating vessel that may or may not contain an agitator bar. The unit combines V shapes to enable continuous mixing. After the initial mix, a lube step can be added, if required, directly into the blender. At this point the product can exit the mixer directly onto a PAT device which senses for content uniformity in real time. At this point, the blended product can be dispensed directly into a tablet press. The blender is fed continuously and runs at a set rate for a set period of time, both determined by the process.

Microencapsulation. Brookwood Pharmaceuticals (Birmingham, AL) has developed both batch and continuous processes for producing microparticles and drug-polymer implants. Its microencapsulation process is conducted two ways, depending on the scale. For laboratory scale during feasibility studies, individual unit operations are segregated with planned stops within the overall process. Larger scales use a semi-batch process with continuous homogenization and mixing.

The patented continuous process begins with an emulsification step, followed by extraction and collection in a sequential manner.The resulting microparticles are then dried and prepared for the filling operation (see Figure 1). Process run time depends on the batch size and water exposure time. "From the time emulsification begins or at the initiation of microdroplet preparation that contain API, polymer, and solvent until we're in the drying portion of the process, we keep water exposure to the product to under 24 hours," explains Bruce Hudson, senior engineer at Brookwood.

Figure 1

Taking ownership

FDA has made it clear that it wants manufacturers to take the lead in moving continuous processing forward.

"What we really need to know and I ask our colleagues in industry is, What can we do in addition in order to facilitate the implementation of these innovative technologies that are currently being used in other industries, continuous manufacturing being one of them? We need to find a way to eliminate any perceived regulatory hurdles and find a way to encourage industry to take ownership and to implement new technologies as they see appropriate," says Nasr.