OR WAIT null SECS
Agnes Shanley is senior editor of Pharmaceutical Technology.
Pharma’s test of continuous manufacturing is starting with oral solid-dosage forms.
FDA’s current good manufacturing practices for finished drugs were first published in 1963 (1), a year before Bob Dylan’s song captured the spirit of change that was to define the decade. GMPs would become law in 1978. At most pharmaceutical facilities, however, little had changed since the 1950s, with the same manufacturing processes and equipment, and the same test-centered approaches to quality assurance.
Decades later, pharma has remained in that same time warp, as Janet Woodcock, director of FDA’s Center for Drug Evaluation and Research (CDER), said at the American Association of Pharmaceutical Scientists’ annual meeting in 2011 (2). She shared predictions that pharmaceutical manufacturers would spend the next 25 years moving to cleaner, flexible, and more efficient continuous manufacturing. Years earlier, FDA had established a framework based on statistics, multivariate analysis, and more advanced process control, to help them reach this goal (3).
Six years into that transformation, change may be more of a trickle than a flood, but it is starting in oral solid-dosage-form (OSD) facilities, as more companies explore alternatives to batch manufacturing. “In manufacturing, nobody in pharma wants to be the first to introduce anything new,” says Dave Di Prospero, director of pharmaceutical process technology at CRB USA and cofounder of the International Society of Pharmaceutical Engineers (ISPE)’s OSD Community of Practice. “But, with FDA approvals of new continuously manufactured drugs by Vertex, and the transition of a legacy therapy from batch to continuous by Johnson & Johnson’s (J&J) Janssen Pharmaceuticals, more companies are getting on board with continuous processing,” he says, “so the ‘first’ is now behind us.” The European Medicines Agency approved Janssen’s continuous line for Prezista (Darunavir) in June 2017, a year after FDA.
Continuous processing is being used at Novartis, which first brought the technology into mainstream conversation with its 10-year, $65-million research program with the Massachusetts Institute of Technology, as well as Pfizer, Merck, and GlaxoSmithKline.
Contract manufacturers are also investing in continuous. Patheon, for instance, completed a continuous manufacturing facility in Greenville, NC earlier this year, appointing former Rutgers post doc and Janssen process engineer Eric Jaycock (4) as director of continuous manufacturing. “I thought they would be the last ones to grab hold, but CMOs want to offer continuous processing as a service, and view it as a competitive advantage,” says Russ Somma, principal of Sommatech Consulting. “Having CMOs in the game allows more manufacturers to dip their toes into continuous processing without risky investment,” he adds.
Generic-drug manufacturers are attracted to continuous, based on its potential to reduce the cost of goods, as well as plant footprint, staffing, raw materials, and solvent requirements, says Bayan Takizawa, an MD and chemical engineer, cofounder, and chief business officer of CONTINUUS Pharma, a spinoff of the MIT-Novartis partnership whose goal is to help more companies understand and harness the benefits of continuous processes. The company offers its Integrated Continuous Manufacturing (ICM) platform (Figure 1), which encompasses continuous production, from raw materials and APIs through finished dosage forms, and was developed through the MIT-Novartis partnership. “Generics margins are significantly lower than those of big pharma, so even if we can reduce cost of goods by 30-50%, that’s a huge impact,” he says.
Pfizer is using continuous on both extremes of the value spectrum, reducing throughput and costs for some of its high-volume generic APIs, but also for high-end products including personalized therapies. “Inherently, you have to understand the process better in order to manufacture continuously than you do in batch, so you have to work harder upfront and invest more during the R&D period with continuous,” says Kevin Nepveux, vice-president of global technology services at Pfizer, “but it pays off in a more reliable process.”
“Agility also becomes more important with personalized therapies,” he says. “And with semicontinuous you can reduce lead time by 50% or more, so you can respond to accelerated approvals or changes in market demand.”
Over the past few years, adoption of continuous processes had increased at a lower than expected rate, says Pamela Docherty, life-sciences industry manager at Siemens. However, she notes that more pharmaceutical companies are now working on continuous processes for legacy products, while more equipment vendors are offering solutions designed for continuous operations. Within the next two years, she says, these changes, plus a growing number of continuously manufactured products in R&D, should result in a second spike in approvals of continuously manufactured drugs that could lead to exponential growth.
By working with academic consortia and pharmaceutical companies, equipment manufacturers have become important partners in the move to continuous processing. “We’ve gotten to where we are now through collaboration involving pharmaceutical companies, equipment vendors, and academia,” says Di Prospero.
GEA Pharma Systems, which entered the continuous area in 2003, established a joint venture with Siemens in 2016, leveraging the SiPAT analytic software and Siemens’ advanced process control platforms, together with its continuous OSD platform, ConsiGma, which is already being used by a number of big pharma companies. The system integrates raw material dosing and blending, wet or melt granulation, drying or cooling, tableting, and coating on one line, with online quality control built in. Other equipment vendors that are active in continuous include L B Bohle, Glatt Process Systems, and IMA, a tablet press manufacturer based in Italy, which is an investor in and strategic partner with CONTINUUS Pharmaceuticals.
Obstacles remain to industry’s adoption of continuous, however. There are some lingering doubts about regulatory support, as well as technical challenges, particularly around real-time release testing (RTRT) (see related article).
On the practical side, there is also a need for equipment designs that would speed cleaning and product changeovers at continuous lines. In some facilities, it can take several weeks to tear down, clean, and set up a continuous line, says Di Prospero.
Novartis and MIT envisioned homogeneous, end-to-end continuous manufacturing in their research, and Novartis is working on an end-to-end continuous facility in Basel, Switzerland. Upstream, continuous processes are being developed for APIs (see related article), and for biopharmaceutical manufacturing.
“Implementing the end-to-end approach is really challenging, from the technical but also from the corporate side, and it’s unlikely that we’ll see more than a handful of companies taking this approach in the next decade or two. It’s simply too front-loaded from a cost perspective, and there’s too much product-specific complexity to be addressed to allow the business case to work out,” says Doug Hausner, associate director of the Engineering Research Center for Structured Organic Particulate Systems (C-SOPS) based at Rutgers University in New Jersey.
C-SOPS (Figure 2) has been around for 10 years now, collaborating with industry and regulatory agencies, and has transferred continuous process knowhow to industrial partners including Janssen, which collaborated with C-SOPs on the continuous direct compression process that it uses at its facility in Puerto Rico (5). Hausner expects continuous to be developed, first, for OSD and then for APIs, but sees limited instances where there will be a clear business case for integrating the two.
Each company’s decision whether or not to adopt continuous processes, and where to use them, will be based largely on its tolerance for risk, says Takizawa. Most companies today are taking a case-by-case strategy.
Adoption will also depend on how opportunities arise, says Jamie Clayton, operations director at Freeman Technology, a UK vendor of powder flow characterization technologies. Many different configurations are possible, involving combinations of continuous, semicontinuous, and semibatch systems, which can all be considered some form of continuous processing. “In some cases, companies may use continuous process equipment for unit operations as discrete elements within a larger overall batch process,” he says.
On the process side, continuous processing has intensified interest in direct compression, a simpler and less energy-intensive OSD process than dry or wet granulation, which involves the addition of liquid, high-shear granulation, and drying, says Hausner.
Continuous eliminates the need for scale up, one of the biggest bottlenecks in drug development, allowing manufacturers to move more nimbly into clinical and commercial production. “With continuous, you don’t scale up; you scale out, by extending your processing time, which can save a lot of money,” says Takizawa.
Eli Lilly, which has also collaborated with C-SOPS, is using direct compression at two identical GMP facilities, one in Puerto Rico and another in Indianapolis, where a third development facility is based. As Hausner explains, the three facilities are identical down to the software, except for a refill system on the commercial production line.
The company’s strategy is designed to leverage information at earlier development stages, and streamline process development. “As soon as they get a new molecule and have enough material available for experimentation and process development, they can do development and have the commercial process figured out by the time Phase II clinical trials begin. This way, they’ve minimized variability within clinical trials, and, if and when the drug is approved, they’ll be ready to go,” says Hausner. The twin GMP plants won two category awards at the 2016 International Society for Pharmaceutical Engineers (ISPE) Facility of the Year Award program (6).
Advances are also being made in continuous wet granulation, aided by research collaboration with equipment manufacturers. GEA has been working with academic consortia and pharmaceutical manufacturers in this area.
Vertex uses continuous wet granulation, while Pfizer uses both continuous direct compression and continuous granulation. The company worked with GEA and other equipment vendors on a proprietary mixing platform that blends multiple ingredients and sizes particles continuously. It includes an automated weigh dispense system, continuous mixing, and continuous granulation allowing users to go from weigh dispense to mixing/granulation to a tablet press. Nepveux and his colleagues plan to add a few more unit operations to Pfizer’s continuous portfolio. “We can select all or parts of the platform to use, depending on the product,” he says, “and it will be the main way that we develop innovative OSD, and transfer legacy products from batch to continuous when it makes sense.”
In one project (6), GEA and Freeman Technology used QbD principles to see how raw material properties influence granule properties in a wet granulation, and how granule properties then determine critical quality attributes in tablets, says Clayton. Using Freeman’s FT 4 powder rheometer and GEA’s ConsiGma 1 high-shear wet granulator, studies showed a direct relationship between bulk flow properties of granules and tablet hardness.
“We’ve done quite a lot of work with dry and wet granulation, to learn how granulate properties can be targeted by adjusting critical process parameters, and how granule properties can then be linked to the critical quality attributes of tablets,” Clayton says. The company is currently marketing Lenterra flow measurement tools, which are process analytical technology (PAT) devices that can measure drag force and wall shear stress to monitor processes in-line or define optimal process parameters or granulation endpoints.
Some of the most important advances in continuous OSD have come on the feed side, says Di Prospero, because it is important, but challenging, to achieve a good feed of ingredients, including API, in a controlled and accurate configuration. Lilly spent a considerable amount of time perfecting this technology, he says. Accuracy is improving as more is understood about how the physical properties of raw materials effect the process (7), and vendors are offering solutions designed for continuous processes, such as Coperion K-tron’s smart refill technology, which uses an algorithm to store and trend weight to speed measurements (8).
GEA, meanwhile, has incorporated better feeding mechanisms in its integrated continuous equipment designs, such as ConsiGma. Miniaturization and close integration have enabled major improvements in equipment design, across the board, says Richard Steiner, business development manager for pharma applications at GEA.
One of the biggest attractions of continuous processing is the reductions in facility size and layout that it permits, says Di Prospero. In a typical batch processing facility, dispensing, granulation, milling, blending, compression, coating, and sub operations are running, each requiring a room and additional support spaces, he says. In a continuous facility, they are linked together and can fit in a single room. “We’ve seen square footages go from 30,000 to 40,000 down to 3000 or 4000 with continuous,” he says. “And with reduced footprint comes reduced utility requirements and less heating, ventilation, and air conditioning (HVAC).
Continuous processes also lend themselves well to containment, which has become a prerequisite since formulations have been using highly active pharmaceutical ingredients, says Somma. “With traditional batch, containment is needed at the dispenser, granulator, and every other transfer point within the process. With continuous, you’ve reduced your need for containment from 15-20 points down to two, says Di Prospero.
Both containment and miniaturization trends come together in the Portable Continuous Miniature and Modular (PCMM), portable, GMP OSD facilities that combine G-Con’s modular production pods and GEA’s continuous processing technology. Pfizer has developed a prototype of the concept, which would allow units to be rapidly deployed where and when needed (9).
Users say that quality by design and process analytical technologies are essential for continuous processes. “You cannot control continuous processes with traditional in-process control sampling and off-line, laboratory-based testing,” says Nepveux.
PAT has improved significantly in the past 5-10 years, he says, to a point where it has become an enabler for continuous, Nepveux says. “We’re getting away from big instruments that you try to plug in through a port.” In the past, he notes, it was a challenge to get a window into a process and be sure that it was truly representative of what was going on in that process. Today, Nepveux says, creative sensing technology avoids this problem.
“There are technically effective and cost-effective PAT sensors that can look at potency, particle size, moisture, blend uniformity, and all the different types of measurements that you might want to take across a process,” he says, “so that part is there now.”
An area where industry remains challenged is in overall continuous process control, notes Di Prospero. “As continuous OSD plants run, there may be situations where, for example, mixer speeds have to change without affecting the endpoints,” he says, so there is a need to control all the processes and have them talk to each other. “The hardware pieces are all there, but the software and the control integration with PAT is still evolving,” he says.
Different companies may take different approaches to control, says Nepveux, with some basing it on residence time, and others on a feedback loop in which real-time results are used to modify key parameters to maintain the same output. “The question then becomes: Are you going to assume that your inputs are fixed, or do you expect them to have reasonable amount of variability, and have processes adapt to that?” he says, noting that Pfizer takes the latter approach.
The other big challenge has been the control systems themselves, not just single loop but multivariate, high-level process control. “You can make a unit operation continuous, but if you cannot control it effectively, then, when you have to connect to other steps, it won’t go very far,” says Nepveux.
However, pharmaceutical manufacturers are showing interest in using advanced process control, or model predictive control, for different interacting components within the line. “They want to go beyond traditional proportional-integral-derivative (PID) control and ensure that the process stays within specifications and remains stable,” says Docherty. “Pharma wasn’t using this type of control all that much in the past, but now some applications require it,” she says.
Pharmaceutical manufacturers are also beginning to recognize the benefit and importance of soft sensors, computer programs that process different measurements and calculate new data points based on the interactions of those measurements, without requiring that data to be physically measured.
“In the beginning, everyone was looking for the best online analytical device, but now that there is a better understanding of process control, people realize that the data required already exist. With these tools, predictive modeling makes a lot more sense,” says Steiner.
One of the biggest concerns for pharma has been regulatory attitudes toward continuous manufacturing. With some modernization programs in the past, such as PAT and QbD, there was high-level support within FDA, but it didn’t always filter down to the troops performing inspections or reviewing products. Some companies may still be concerned about this happening with continuous says Docherty.
“It could be an issue if a lot of people were suddenly moving to continuous,” says Hausner. “But the adoption rate has been gradual.” Besides, he notes, FDA has recruited more engineers (including some C-SOP’s graduates) and revamped its inspection program so that inspectors either specialize in food or pharma, but don’t have to do both.
Clearly, FDA is investing more resources in better understanding continuous processing, says Hausner, and, as more successful results have been demonstrated, the agency has gone from merely accepting the technology to actively supporting, even advocating for it, he says.
At the end of 2016, CONTINUUS began work with FDA, constructing an end-to-end continuous line that will be used to study and better understand phenomena that are critical for the regulatory review of continuous systems, such as RTRT and traceability, says Takizawa. C-SOPS has helped train FDA staff in continuous concepts and methods, says Hausner.
Meanwhile, EMA’s approval of Janssen’s continuous line in Puerto Rico may inspire more manufacturers. Regulators have already invited companies to discuss issues with them before implementing continuous and including it in a new drug application or a post-approval change request, Hausner says.
Because pharma is still at a relatively early stage in its exploration of batch alternatives, it faces a “chicken and egg” situation, says Hausner. Companies want more case studies to convince management to invest in continuous, while regulators want to see more applications before they issue specific guidance documents, he says, noting that lack of guidance can then detersome companies from filing applications with continuous.
Recently, the US Pharmacopeial Convention (USP) has become more involved in continuous. The group held a meeting on the subject in June and previously released a video on the topic on YouTube (10). Its involvement could help convince more companies to evaluate continuous more closely. In the meantime, a growing number of implementations suggest that more manufacturers are questioning the status quo, nudging OSD manufacturing out of its time warp and toward more modern approaches.
1. P. Melamud, “A Brief History of USFDA GMPs,” Presentation, ISPE’s NJ Chapter Day, June 17, 2009.
2. S. Chatterjee, “FDA Perspectives on Continuous Manufacturing,” Presentation, IFPAC General Meeting, January 2012.
3. FDA, “Innovation and Continuous Improvement in Pharmaceutical Manufacturing,” Report, April 2004.
4. Patheon, “Patheon’s Eric Jaycock Discusses Continuous Manufacturing in Pharma,” Video, March 28, 2017.
5. Freeman Technology, “A QbD Approach to Continuous Tablet Manufacture, Application Note, freemantech.com.
6. ISPE, 2017 ISPE Facility of the Year Award Winners Announced, Press Release, February 6, 2017.
7. C. Blackshields and A. Crean, “Continuous Powder Feeding for Pharmaceutical Sold Dosage Form Manufacturing: A Short Review,” Report, June 23, 2017.
8. S. Nowak, “Improving Feeder Performance in Continuous Pharmaceutical Operations,” Pharm. Tech., 40 (10), 68-73.
9. J. Markarian, “Paving the Way for Continuous Manufacturing,” Pharmaceutical Technology Equipment and Processing Report, April 19, 2017.
10. USP, “Pharmaceutical Continuous Manufacturing Technology and Applications,” Video, March 14, 2017,
Vol. 41, No. 7
When referring to this article, please cite it as A. Shanley, “21st-Century OSDs: Times, They Are a Changin’,” Pharmaceutical Technology 41 (7) 2017.