Actions are speaking loudly
The FDA is making meaningful internal changes so that pharmaceutical manufacturers can move into the 21st century
By Sheila Galatowitsch
Working at lightning speed, the U.S. Food and Drug Administration (FDA) has made good on its promise from August 2002 to re-assess current good manufacturing practices (cGMPs) in pharmaceutical manufacturing.
The short list of its achievements includes a final guidance on electronic records and signatures, accomplished in just one year; draft guidances on aseptic processing, process analytical technology (PAT), dispute resolution and comparability protocols; the creation of a "pharmaceutical inspectorate"; and international outreach.
What's driving all this activity? The agency's awareness that its own regulatory inflexibilities and inconsistencies are keeping U.S. pharmaceutical manufacturers from adopting advanced technologies, such as isolators and online testing with PAT. These innovations would ensure a steady supply of the most critical drugs as well as save on manufacturing costs and improve product quality.
The FDA started running with the cGMP initiative as soon as it was announced 17 months ago. "We've gotten used to it now, but when the first six-month status report was issued, we were shocked at how much they had accomplished," says Gerry Migliaccio, vice president of global quality operations for Pfizer Inc. (New York, N.Y.). "We have now come to expect that they are serious about this effort. It's not just another program. They have staffed it properly and mobilized the right people."
Industry, too, has dedicated senior personnel to the initiative. Migliaccio and his peers at companies such as Merck and AstraZeneca carve out significant portions of their workday on behalf of the undertaking. Association working groups are coordinating white papers and guidance comment to avoid duplicating effort.
Stakeholders are optimistic that the agency, with its new emphasis on quality, risk and science, is moving from a "change is bad" to a "change is good" mindset, says Migliaccio. The joint effort between the FDA and industry, he says, is facilitating innovation that will "provide us with significantly more knowledge about our processes, and while we put out a very good quality product now, we will be improving the reliability of our operations."
Electronic record-keeping redefined
Much work remains to be done, but life is starting to change for manufacturers, especially now that the FDA has completed its final guidance on Part 11 of Title 21 of the Code of Federal Regulations, which provides criteria for the way electronic data is maintained and submitted to the government.
The 1997 regulation was intended to permit the widest possible use of electronic records and signatures, but because it was interpreted so broadly by the agency, it had the opposite effect. Over the past seven years, Part 11 requirements for validation, audit trails, record retention, record copying and legacy systems have, in some cases, served to discourage technological advances.
This realization finally hit home to regulators when manufacturers said they might not use PAT, which the agency believes is critical to public health, because "the data that would be collected under Part 11 would be too hard to archive," says Paul D'Eramo, International Society for Pharmaceutical Engineering (ISPE) vice chairman and a former FDA inspector and manager of technical operations.
PAT would improve the quality of manufacturing processes with continuous monitoring of entire process lots. For example, it uses new technologies, such as laser and infrared sensors, that can analyze potency in every drug product coming down the line.
"The problem is the amount of data collected in eight million data points, when you had only 10 data points before," says D'Eramo, who is also executive director for worldwide policy and compliance management at Johnson & Johnson (New Brunswick, N.J.). "It would cost manufacturers a fortune archiving electronic data under Part 11, because it would be considered raw data, and all raw data has to be kept."
Redefining Part 11 became the agency's first priority and first accomplishment in its cGMP initiative. In the final guidance issued last August, regulators reduced Part 11's scope and applied the concepts of risk management to the use of electronic data, giving manufacturers more flexibility in judging the importance of electronic records to their operations.
"Now a company simply has to comply with predicate rules and justify with a risk-analysis document why they need or don't need electronic data in all cases," says D'Eramo. Under the new guidance, manufacturers are beginning to explore how they can use resources better and save on costs, especially when remediating equipment.
The Part 11 guidance is a "tremendous step forward," says Migliaccio, and sets the stage for introduction of PAT. And even though the agency is still working on its PAT guidance, several manufacturers have filed for approval of PAT methods.
The PAT draft guidance describes a regulatory framework to encourage the voluntary design, development and implementation of innovation manufacturing and quality assurance technologies. The framework has two general components:
- A set of scientific principles and tools supporting innovation;
- A new regulatory strategy for accommodating innovation, such as joint training and certification of PAT review and inspection staff.
This new strategy "should alleviate any concern among manufacturers that introducing new technologies will result in a regulatory impasse," the agency wrote in its September 2003 Progress Report.
Updating aseptic processing
Another major highlight of the cGMP initiative is a draft guidance on sterile drug products produced by aseptic processing— a long-awaited update to the 1987 guidance document.
The FDA said it designed the draft guidance to underscore the advantages that automation and isolation technologies offer in protecting the exposed sterile drug product during its aseptic manufacture. It also tried to clarify regulatory expectations to reduce the incidents of manufacturing problems with this class of pharmaceuticals, which are often a major cause of drug shortages.
The draft provides recommendations on how to build quality into products using science-based facility, equipment and systems design, and it includes two risk-related themes. The first ensures that operational and raw material inputs are predictable through adequate quality control and quality assurance, and the second ensures reliable and robust product protection through adequate design and control.
Although hailed as a significant improvement upon both the 17-year-old document and a FDA 2002 concept paper on the topic, the guidance document nevertheless elicited criticism for its perceived flaws. The PDA's (Bethesda, Md.) 20-member Aseptic Processing Task Force, co-chaired by Richard M. Johnson of Abbott Laboratories and Martin VanTrieste of VanTrieste & Associates, prepared comments calling for additional clarity in the document.
The task force suggested removing certain items that are covered in other guidances, deleting reference to products and processes other than aseptic processing, and asked for clarification that the document does not apply to terminally sterilized products. It also suggested the use of internationally standardized units throughout the draft guidance, which when finalized, will replace the 1987 guidance.
The FDA received feedback from numerous groups and individuals, including Russell E. Madsen, president of the consulting company The Williamsburg Group (Gaithersburg, Md.) and a former science and technology executive at the PDA.
"In general, it is a good document and improves on the '87 guidance in several ways; but in terms of its scope, it goes too far," says Madsen. "For example, it contains things not directly related to aseptic processing, like sterilization qualification and validation, equipment controls and instrument calibration, and sterility testing. These should be the subjects of separate guidance documents."
Madsen adds that the draft guidance also has an "implied expectation of sterility for aseptic processing environments that really is inappropriate. These are very, very clean, decontaminated environments, but they shouldn't be considered to be sterile any more than a general cleanroom environment."
Another concern is the draft's unrealistic expectations about the ability to accurately quantify microbial contamination. "These measurements at best have a +/- 20 percent accuracy," says Madsen, "yet this document talks about being able to distinguish between one and two microorganisms—and that's just not possible."
Despite these criticisms, the draft is better in many respects than the 1987 guidance, says Madsen, and that's due to greater specificity on certain compliance issues, such as media fill acceptance criteria. "The downside is it stifles more advanced aseptic processing technologies, such as isolators and blow-fill-seal, and adds some unrealistic expectations in respect to environmental monitoring and controls that are not necessary in some of these high-quality environments," Madsen adds. (For a detailed look at comments submitted, see adjacent story, "Aseptic processing: Making a good document better.")
Resolving disputes and more
Madsen joins other industry partisans with praise for FDA's work on a dispute resolution process. Previous to the recent draft guidance, there was no formal process to resolve scientific disagreements between the agency and industry. Disputes would frequently drag on for weeks or months without resolution. The government's plan for a dispute resolution program will launch a new era in agency/industry communication.
And with an eye toward improving consistency in inspections, the agency created a pharmaceutical inspectorate (PI) that has 50 investigators who are highly trained in pharmaceutical manufacturing. The PI will focus on conducting human drug quality inspections of prescription drug manufacturers and other complex or high-risk pharmaceutical operations.
ISPE has already sponsored several one-week training courses, taught by instructors from the pharmaceutical industry and hosted by individual drug companies. At the first course hosted by Bristol-Myers Squibb (BMS), experts in different manufacturing areas gave lectures to review chemists, who then toured the BMS facility. A second course was held at the Johnson & Johnson research fab, and a third is scheduled for this month at Roche.
In other efforts to improve consistency, the agency has increased the use of product specialists to coordinate submission reviews and cGMP inspections, and has revised the regulatory procedure it uses to issue warning letters.
Still more highlights of the agency's work are detailed in the adjacent story, "Moving mountains at the FDA" (page 13). This year, the agency and industry will take the leap from talking about these concepts to putting them into action. "Operationalizing is the hard part," says Migliaccio. Now the real work lies ahead.
SHEILA GALATOWITSCH, a special correspondent to CleanRooms, is based in Denver, Colo. She can be reached at: firstname.lastname@example.org
Moving mountains at the FDA
Industry observers are still shaking their heads at the speed and intensity of the FDA's work on its cGMP initiative.
"I didn't know a government agency could move that fast, but it's refreshing to see that when they put their minds to something, they can make it happen," says Russell E. Madsen, president of the consulting company The Williamsburg Group (Gaithersburg, Md.), and a former science and technology executive at the PDA.
Last September, one year after announcing the pharmaceutical cGMP initiative, the FDA released its second Progress Report and Implementation Plan. In 11 pages, the agency summed up a year's worth of work, beginning with the final guidance on Part 11, Electronic Records and Signatures.
In addition to the draft guidances on aseptic processing and PAT (covered in this month's Special Report), the agency drafted guidances on dispute resolution and comparability protocols.
The dispute resolution document targets industry/agency disputes about scientific and technical issues relating to cGMP regulations. In January, the agency launched a 12-month domestic pilot program consistent with the guidance document, which encourages open, prompt discussion and dispute resolution.
When finalized in early 2005, the dispute resolution guidance will describe procedures for raising disputes to the Office of Regulatory Affairs and agency centers and for requesting reviews by a dispute resolution panel.
The comparability protocols document will let manufacturers of protein-based human and veterinary drug products and biological products implement some changes without submission of a prior approval supplement. It describes recommendations for preparing and using pre-defined change evaluation plans-generally referred to as comparability protocols-to implement postapproval manufacturing changes.
This year, the agency plans to finalize a similar draft guidance for nonprotein human and veterinary drugs, and look for additional ways to reduce application submission and filing requirements. Throughout the year, it will identify opportunities for managing manufacturing changes without the need for prior FDA review or approval.
For texts of the draft guidances and more information on the cGMP initiative, visit www.fda.gov/
cder/gmp/index.htm. Here are brief updates of other initiative components:
- The Preapproval Inspection Compliance Program now uses a risk-based approach to determine whether a preapproval inspection is warranted. FDA has revised and reduced the number of mandatory inspection categories that would automatically prompt an inspection when a drug application or supplement is reviewed by the agency. This change gives the agency additional resources for high-risk preapproval and postapproval cGMP inspections.
- The agency's Manufacturing Science working group is working with the PAT team to ensure that existing application review and cGMP programs are based on sound, state-of-the-art scientific and engineering knowledge. And by focusing on process understanding, the agency wants to facilitate risk-based regulatory decisions and innovation, as well as the use of appropriate risk identification, management and control methodologies. To that end, the agency is holding workshops domestically and overseas to discuss innovative scientific risk-based approaches.
- Several collaborations have also been formed, including one with the National Science Foundation's Center for Pharmaceutical Processing Research, in support of PAT and other cGMP initiatives. A cooperative research and development agreement with Pfizer will research chemical imaging applications in pharmaceutical manufacturing and quality assurance.
- In October, the agency will pilot a quantitative, risk-based site-selection model for cGMP inspections. The model, which includes risk factors such as compliance history and manufacturing processes, will help regulators predict where inspections are most likely to achieve the greatest public health impact. Professors at Georgetown University's McDonough School of Business (Washington, D.C.) and Washington University's Olin School of Business (St. Louis, Mo.) are working with the FDA to further refine the model.
- The agency plans to revise the sterile products compliance program for Center for Drug Evaluation and Research products to incorporate risk-based approaches. Other compliance programs targeted for risk-based approaches include the preapproval inspection program and active pharmaceutical ingredients program.
- Also under development is an integrated agency-wide, risk-based quality management system, which will feature educational guidance documents to encourage the use of quality system principles. The first guidance may be released for comment by August.
- Finally, the agency is cooperating with international regulatory authorities. Last year, it attended International Conference on Harmonisation meetings, where expert working groups were created to incorporate risk and quality concepts into a product's lifecycle, and develop a risk management framework for pharmaceutical manufacturing.— SG
Aseptic processing: Making a good document better
The industry consensus is that, in many ways, the aseptic processing draft guidance is an improvement over the 1987 guidance. For example, it provides manufacturers with specific guidance on media fill acceptance criteria where the old document did not.
In the case of isolators, however, the draft places constraints on manufacturers that may inhibit broad-based adoption of the technology. "They are asking people to do things for isolation technology that aren't routinely done for conventional cleanrooms or manned aseptic processing areas, such as their emphasis on glove integrity and isolator sterility over and above what is expected in conventional environments," says Russell Madsen, who served on the PDA's task force that submitted comments to the agency. He also submitted an additional 25 pages of comments in his role as private consultant.
When it errs, the draft steers away from one of the agency's key goals: to focus resources on high-risk processes. "This draft adds requirements to aseptic processing when there's no evidence to suggest a public health concern," says Madsen.
For example, the draft suggests that cleanrooms used for aseptic processing should be evaluated under both as-built and static conditions, but that the classification of the cleanroom should be conducted under dynamic conditions with personnel present, equipment in place and operations ongoing.
In its response, PDA said that this position has the potential to "both require unnecessary cleanroom evaluation and to further blur the distinction between classification of cleanrooms and monitoring of their contamination-control performance." PDA asserted that cleanroom classification should be performed primarily under static or as-built conditions as defined in ISO 14644, and that evaluation of the dynamic performance of a cleanroom should be left to the monitoring program.
The lack of harmonization with this draft and international standards is another area of concern for Madsen: "These groups on both sides of the Atlantic need to get together and decide how best to integrate these documents to come up with something everybody can live with."
The agency, he says, could have made the document better by simply eliminating old sections on sterility testing and sterilization from the previous guidance, and concentrating on aseptic processing. Still, Madsen's not convinced a guidance document is the appropriate mechanism to regulate aseptic processing.
A better way to accommodate the many different varieties of aseptic processing is for the FDA to develop a general template, which manufacturers could use to describe and justify their procedures according to scientific, risk-management and quality systems principles. The agency could then use the specific approved applications to hold manufacturers accountable to their own standards.
Given the time it has already spent on the draft guidance, the agency is unlikely to adopt such a template approach. At press time, the agency had not announced whether it would issue a second draft for comment or publish a final document later this year. — SG