Process analytical technology is intended to make pharmaceutical manufacturing more efficient, but are the right people aware of PAT’s benefits?
By Carolyn Mathas
“The pharmaceutical industry has a little secret: its manufacturing efficiency lags behind that of potato chips and laundry detergent,” said a front-page Wall Street Journal article in September 2003.1 What might have been a little secret grew to big news and further, toward formulating a solution with a great impact on the industry.
In the early 2000s, pharmaceutical regulation at the FDA meant dealing with companies in two ways: issuing warning letters or placing them under consent decree. Drug shortages because of manufacturing difficulties were commonplace and recalls were rising. In 2002 there were 354 drug recalls, up from 176 in 1998. The numerical estimates of the industry’s process efficiency ranged between 3 percent and 15 percent, compared with the semiconductor industry at 99 percent or better.
Yet, there are many similarities between the two industries: batch processes drive both; both have extremely high-value products offset by short life spans; and both industries are highly competitive. Even so, technologies used or lessons learned in the semiconductor industry are not even on the radar screen for most in the pharmaceutical industry.
Referring to the low process efficiency rates, Ajaz Hussain, deputy director of the Office of Pharmaceutical Science in CDER, FDA, says, “That was the sound of the guillotine…With a focus on consent decrees, were we really improving the quality of the product that goes to the patient, or were we simply improving the quality of the documentation for FDA inspections?” The question was finally enough to begin to move towards change.
The perils of manufacturing
What was the root cause of the inefficiencies, and what could the FDA do regarding the manufacturing challenges of the industry? The basic challenge centered on the fact that manufacturing processes in the pharmaceutical industry were obsolete and, up to that date, companies were rewarded for keeping them obsolete.
Manufacturing’s primary emphasis at the time was to get product out the door, which discouraged changes to existing processes. Simultaneously, FDA’s rapid evolution from a sole emphasis on basic procedures and recordkeeping to a focus on good manufacturing practices (cGMP) initially met with industry skepticism.
“Manufacturing processes in the pharmaceutical industry are based on the old way of doing things,” says Dr. Hussain. “So we made a proposal to the FDA Advisory Committee for Pharmaceutical Science in July 2001 that we would like to examine this issue further and would like to propose a modern approach to pharmaceutical manufacturing.” A key meeting that would form a foundation for change was held in November 2001 when a small FDA team went to the highest advisory committee at the FDA, inviting pharmaceutical companies, universities and others, to delve into their perspectives on the current state of manufacturing.
“It was a landmark meeting,” says Dr. Hussain. “We learned so much.” According to Dr. Hussain, Pfizer’s Senior VP of Manufacturing Norman Winiskill acknowledged that Pfizer had a policy of “don’t use and don’t tell” when it came to modern manufacturing methods. Don’t use, because they were afraid the FDA might not approve them and regulatory uncertainty was high. Instead, they kept with old methods. “Or, they may use modern methods to manufacture their product,” Dr. Hussain speculates, “but they won’t tell the FDA. They use them and then provide the same old ‘stuff’ that the FDA wants because of the regulatory uncertainty,” he adds.
Also surprising, G. K. Raju of MIT presented a benchmarking study that showed that manufacturing efficiency in the pharmaceutical industry, even for making tablets and capsules, was incredibly slow. “It was taking an average of 60 to 90 days to manufacture and release a batch of tablets that should be made in one 8-hour shift. Much of that time was spent on deviations that needed to be investigated and on testing,” explains Dr. Hussain.
According to Dr. Hussain, the remarkable revelations continued at a subsequent meeting when Ray Scherzer, senior VP of manufacturing for GlaxoSmithKline, reported that the cost of manufacturing was three times the cost of R&D. He claimed that inefficiencies and waste made manufacturing the second highest cost item after marketing, much higher than R&D.
Dr. Hussain says that Scherzer further observed that manufacturing in this industry wasn’t working and that quality by design was what was needed. Kicking and screaming, the new process analytical technology (PAT) initiative was officially born.
PAT entered a bleak scene. A worldwide slowdown in new drug and biologic applications to the FDA began in 2000. Still healthy were the costs of bringing new medicines to market, a range estimated by the FDA to be as high as $0.8 billion to $1.7 billion, an increase of 55 percent over the past five years. The high price tag reflected a concentration on drugs with commercial viability, while other drugs that needed to be developed and those for third-world countries continued to be shelved.
Oddly, it isn’t a lack of technology discovery, but delivery that appears to be causing problems. As long as archaic manufacturing methods are used and high failure rates abound, the latest breakthroughs in discovery can’t make it to market. For change to take place, process development must modernize and be efficient and predictable to deliver safe drugs in a timely fashion.
It was as much a matter of who was in charge as it was the outmoded processes. Quality performance had always been established in the laboratory and it was the lab staff who tested products, approving or rejecting them without sharing the data that they generated. As Dr. Hussain points out, “Quality cannot be tested into pharmaceutical products. Rather, it results from a quality-oriented process design…Without specific on-stream analytical instruments, final product quality is neither predictable nor controllable.”
The FDA was not the only organization shocked by the low efficiency numbers. In order to verify what was actually true, a number of companies, such as Merck and Pfizer, were visited. These initial meetings were quite revealing. At one of the first companies where Dr. Hussain presented G.K. Raju’s data, hands flew up immediately after the presentation and Dr. Hussain was told that he didn’t know what he was talking about.
“Luckily for me, the company I was visiting was part of Dr. Raju’s benchmark study,” says Dr. Hussain. “I told them that I could not answer the question as to whether their efficiency was this low or not, but that I’d let the VP of manufacturing-who happened to be in the front row-answer the question. They kept saying, ‘How can this be?’ And the VP of manufacturing verified the numbers. I had their own people reconfirm those numbers.”
Within the pharmaceutical industry, different organizations, such as R&D, manufacturing, and those responsible for regulatory issues, do not communicate effectively and the state of manufacturing efficiency was at the time unknown to them. In this case, the question came from regulatory affairs and not R&D.
Outreach began with a number of workshops held by associations, the FDA, and academia-including schools of engineering and pharmacy in the United States, Europe and Japan. The outreach provided champions within both academia and the pharmaceutical industry.
Goals of the FDA
Dr. Janet Woodcock, FDA deputy commissioner of operations, had made presentations to the FDA Science Board indicating that things needed to change. Armed with data from academia, companies, and its own findings, the FDA began to move towards a new system based on process analytic technology. What PAT added to the mix were an engineering perspective and scientific understanding. A policy was established whereby the FDA would be conducive to innovation and continuous improvement that would not be imposed, but would be voluntary. The FDA would co-create change rather than mandate it.
To support the policy, the FDA created guidance and supported that guidance with training and certification provided by a core group of FDA staff. So instead of requiring manufacturers to use a particular technology or method, the FDA identified the key issues that they should be considering and created a framework guidance. A team was formed that was open and that communicated. “We had a wonderful program of outreach and an opportunity for companies to come with their proposals,” says Dr. Hussain. “That’s how it went from our being shocked at what existed to having companies say, ‘All right, let’s do it.’”
Unveiled in August 2003, the PAT initiative is part of the FDA’s 21st century current good manufacturing practices (cGMPs). Under PAT, quality and performance are attained and assured by designing effective and efficient manufacturing processes. By understanding how process factors impact end-product performance, product specifications through end results are looked at completely differently. The result of a focus on processes is predictable and efficient manufacturing, with a corresponding continuous improvement in product quality.
PAT is a framework for designing, analyzing and controlling manufacturing through a series of timely measurements taken throughout the manufacturing processes. Both raw and in-process materials are monitored and measured as well. The ultimate goal is final product quality, yet additional money-saving benefits include faster new product development, shortened manufacturing cycles, improved yields, less waste, and fewer product recalls.
To encourage PAT adoption, the FDA is streamlining the regulatory process for adopting new technologies within the manufacturing environment and harmonizing regulatory expectations. This allows PAT companies to take advantage of technology advances that they were hesitant to implement in the past. Many measurement tools designed to eliminate process variables meet specific and predetermined incremental goals and are used to guarantee results. In addition, companies themselves establish and control the variables that are most critical to their own end product. Where the measurement controls are placed is established based on preventing problems and errors rather than catching them after they occur (see Fig. 1).
Figure 1. PAT-oriented products, such as this particle counter with 21CFR Part 11-compliant data download capabilities from Particle Measuring Systems, are making it easier to comply with regulatory requirements.
Risk management tools used under PAT include data acquisition and analysis tools, process analyzers, process and endpoint monitoring and control tools, and knowledge management tools. Online data analysis provides analysis of events, including time of occurrence, duration and magnitude, cause and effect analysis that tracks problems to their source, impact analysis that compares the effects of change, and trend analysis, which is a historical record of equipment performance data. Equally important is the fact that communication across disciplines-with chemists talking to engineers and pharmacists-is developing. Only when all these groups are brought together will it be possible to have a successful PAT-based environment.
Still, adopting PAT is not without concern within the manufacturing community. Of utmost importance is convincing companies that using PAT will not increase risks and delays. Although there is a competitive advantage to doing so, implementing PAT requires articulating its possibilities, how it fits existing corporate goals, and how it will improve productivity and yield to impact the bottom line. Finally, it’s the sale of the concept to management, who are still unaware of PAT, that seems to hold up its implementation.
PAT is intended to benefit both the FDA and the pharmaceutical companies implementing it. With PAT, there is a connection at the FDA between review, inspection, and compliance. According to Dr. Hussain, “There was such a big divide within the FDA organizational structure that communication and collaboration between different organizations was at a minimum.”
For companies, benefits include the mind shift that has taken place in process validation. Much of the industry has grown up with the concept that if they have approval, that approval must be validated. Companies have their best people validate that they use the same amount of raw materials-keep their fingers crossed-and perform three batches to validate and they’re finished. Under PAT, that concept is challenged. Says Dr. Hussain, “That’s not really what process validation is supposed to be. It’s a life cycle approach. If you think you’re going to just do three batches and you call yourself validated, think again.” Validation based on continual quality assurance, where processes are continually monitored, evaluated and adjusted under PAT, may be achieved with less time and resources. Implementing good manufacturing techniques will always reduce a plant’s total cost of production.
Pharmaceutical companies using PAT experience easier transfer of documentation and data, easier regulatory adherence and compliance, lower validation expense, and cursory FDA inspections. Making their way through the FDA maze faster is a reality. Data acquisition methods inherent under PAT include continuous process records, data archives, data retrieval and display-all complying to FDA electronic data record guidelines.
By putting process controls in place and making adjustments based on analysis, the detection and mitigation of problems and process deviations are less costly and more likely to justify modifying regulatory paths and post-approval changes. Online monitoring enables users to analyze production and respond, all in real time.
Adoption so far
“For a lot of companies, nothing has changed; they are still waiting on the fence,” says Dr. Hussain. “There are only a few companies that have…rapidly changed their complete organizational structure.” Dr. Hussain reports that GlaxoSmithKline completely revamped their systems to be PAT-based, but that very few companies have really moved in this direction and some just don’t know how to proceed.
The FDA is comfortable that it has done what it can. It does see step-by-step transformation in the adoption of quality by design. “I hope the companies who have adopted and moved forward on PAT will be able to show a business case because that is what will continue this program,” suggests Dr. Hussain. “I keep saying that there is not an original thought in the PAT guide. [It’s] a 60-year-old concept that we have brought from the semiconductor industry and from all tested and proven methodologies.” The PAT guidance, Dr. Hussain says, is nothing but a conduit or a connection between pharmaceutical quality, modern quality systems, and significant things that have happened in quality circles outside the pharmaceutical industry.
To implement PAT requires that a company perform an appraisal to see where PAT can benefit most. Once an area or product for implementation is identified, a discovery phase involves re-evaluation of practices, processes, and inspection and validation methods. In the analysis phase, product findings, exception history, and manufacturing and quality data for each product is studied. Implementation, however, requires properly developed processes, training in multiple disciplines-including process control engineering, process analytical chemistry, instrumentation, and metrology-and qualified equipment.
Restructuring production lines is expensive, yet the savings from efficient resource use, waste reduction, rapid product approvals and the elimination of product recalls far outweigh PAT implementation costs.
One of the challenges noted by the FDA is the tendency for companies to want the FDA to tell them what to do each step of the way. “Our guidance on PAT has been a broad framework guidance,” states Dr. Hussain. He notes that many companies want specific, step-by-step instructions, but that innovation is ultimately up to them. “The leadership for this has to come from industry,” he says.
In December 2003, the FDA took steps to advance the development of science-based standards for process quality by joining a group of stakeholders from the pharmaceutical community to establish the ASTM International Technical Committee E55 on the pharmaceutical application of PAT. The committee addresses issues related to process control, design and performance, as well as quality acceptance/assurance for the pharmaceutical manufacturing industry.
Having stakeholders with multidisciplinary expertise allows consensus development with a focus on manufacturing process and adds an engineering dimension traditionally missing. Committee E55 is organized with a main executive committee (E55.90) and subcommittees (E55.01, E55.02, and E55.91) tasked with specific topics concerning PAT’s application (see Fig. 2).
Figure 2. The ASTM International Technical Committee E55 addresses issues related to process control, design and performance, as well as quality acceptance/assurance for the pharmaceutical manufacturing industry.
The executive committee consists of elected officials and appointed members-at-large. It oversees the general direction and administration of the committee. The committees are involved with developing strategic documents that provide a framework, including standard practice for process understanding and risk management. Other committees address terminology, measurement system fitness for use, and data management.
According to the ASTM, “The scope of the committee shall be development of standardized nomenclature and definitions of terms, recommended practices, guides, test methods, specifications, and performance standards for pharmaceutical application of process analytical technology. The committee will encourage research in this field and sponsor symposia, workshops and publications to facilitate the development of such standards. The committee will promote liaison with other ASTM committees and other organizations with mutual interests.”2
Collaboration with ASTM provides industry the opportunity to:
- Learn from other industrial sectors, such as petrochemicals, where process analyzers, statistical principles, and risk management have been in use for a number of years.
- Focus efforts on “process” and bring a much needed (pharmaceutical) engineering dimension.
- Work towards international consensus to support other FDA activities such as the International Conference on Harmonization (ICH).
- Involve all stakeholders and multidisciplinary expertise, as the FDA does not have sufficient resources to do this.
What the market says
Over the past few years, several companies have made PAT-related announcements, and the momentum is continuing. For example, PerkinElmer (Wellesley, Mass.) and Procognia (Berkshire, U.K.) announced in a printed statement their method for analyzing the glycosylation of proteins. They specifically target the PAT initiative. “U-c fingerprint technology requires no sample purification or pretreatment, and allows the quantitative analysis of 20 samples in approximately three hours. In addition, glycosylation analysis throughout the therapeutic protein discovery, development and manufacturing process will bring biopharmaceutical companies closer to satisfying the FDA’s PAT initiative. This technology is focused on reducing the time and cost of producing higher quality biopharmaceuticals,” said the company.
BOC Edwards (Windlesham, U.K.) introduced a check-weighing system, emphasizing its use during the manufacturing process. According to a press release from the company, “Any problems with the process are identified virtually immediately and corrected via process feedback to the BOC Edwards filling equipment-a real-time approach that lies at the heart of the FDA’s recently unveiled PAT initiative.”
Axsun Technologies (Billerica, Mass.) launched a new IntegraSpec XP NIR Micro-Optical Spectrometer Platform for PAT applications. Petros Kotidis, Axsun’s vice president of marketing and business development, said in a news release, “We designed the IntegraSpec XP family to be the critical engine of a complete PAT solution, providing real-time feedback for continuous quality monitoring. Working with solutions partners, including analytical instrument manufacturers, system integrations and process control product and service suppliers, our goal is to help manufacturers serving the pharmaceutical industry meet their customers’ PAT objectives and realize dramatic improvements in utilization, throughput, yield, and quality. Miniaturized and extremely rugged IntegraSpec XP spectrometers can be configured for NEMA compliance and Hazardous Zone 1 operation, enabling pharmaceutical manufacturers to see inside, understand and ultimately control their critical processes.”
Reportedly, GlaxoSmithKline (Research Triangle Park, N.C.) was the first company to obtain approval to release a prescription product to market using a rapid detection technology under the FDA program.
In April 2004, Bosch Packaging (Minneapolis, Minn.) announced 100 percent noncontact check-weighing (NCCW) inspection equipment as part of a complete line solution for integration into its pharmaceutical filling line systems. According to the company, the equipment offers increased quality control and efficiency in accordance with the FDA’s PAT initiative.
On April 26, 2005, Bruker Optics (Coventry, U.K.) and Patterson-Kelley (East Stroudsburg, Pa.) announced a partnership for dedicated pharmaceutical manufacturing solutions. Dan Klevisha, VP of sales and marketing for Bruker Optics, announced in a press statement, “We are expecting to bring a technology know-how and complete PAT solution to our key pharmaceutical manufacturing customers. At Bruker Optics, we are constantly improving our Fourier Transform and dispersive analytical instruments, developing new technologies to bring dedicated solutions for pharmaceutical industry. Our new noncontact LANCER NIR analyzer can be mounted onto Patterson-Kelley blending equipment to wirelessly transmit PAT feedback on the content uniformity, blending uniformity, moisture and other important process criteria.”
Although there is movement towards PAT adoption, Dr. Hussain expresses concern. PAT doesn’t mean converting to online analytical sensing. PAT is a means to bring efficiency of manufacturing in line with that of other industries. “If I look at the semiconductor and automotive sectors where manufacturing efficiency is so important to the bottom line of the business,” says Dr. Hussain, “cost savings [potentially] exist in the pharmaceutical sector by improving efficiency. That is apparent to most of us, and it’s also apparent to the engineers in companies. But these folks have a tough time reaching the CEOs of companies, who often have no clue what this is and what it does. That is the biggest surprise to me.” III
- Aboud, Leila, and Scott Henry. “New prescription for drug makers: Update the plants,” Wall Street Journal, September 12, 2003.
- American Society for Testing and Materials (ASTM) Web site, http://www.astm.org. Accessed July 2005