FDA's DRUG QUALITY INITIATIVE

FDA's DRUG QUALITY INTIATIVE hit its one year anniversary at full stride with new guidance documents, outside collaborations and review and inspection program changes emerging. In early September, the agency released draft guidances covering aseptic processing, dispute resolution, comparability protocols and process analytical technology, along with a final version of a guidance on Part 11. Having moved quickly through the FDA pipeline, the guidances are designed to push industry to take advantage of the advancing technology and quality science while reducing potential regulatory barriers. A specially-trained cadre of pharmaceutical investigators is among inspection program changes the agency is making to create more field/center consistency and help allay industry's regulatory concerns about making technology upgrades. [A discussion by CDER Office of Compliance Director David Horowitz of progress made during the first year of the drug quality initiative is included in the box on pp. 17-24.]

Initiative's Speed And Scope Reflect Level of FDA Commitment

FDA's drug quality initiative is moving forward on a variety of fronts in the effort to better harness the power of advancing technology and quality science.

In early September, at the initiative's one year mark, FDA issued a progress report announcing the release of four separate draft guidance documents for public comment covering aseptic processing, dispute resolution, comparability protocols and process analytical technology, respectively, along with the final version of a guidance on Part 11. The guidances were shepherded by working groups operating under the initiative's steering committee.

The progress report also highlights collaborations FDA has entered into during 2003 with: l Pfizer to research chemical imaging applications in manufacturing and control l the National Science Foundation's Center for Pharmaceutical Processing Research to explore new manufacturing analytical technologies l academics at Georgetown and Washington University business schools to study factors leading to superior manufacturing performance, and l the International Conference on Harmonization (ICH) to develop a harmonized pharmaceutical quality system applicable across the life cycle of a product.

Inspection-related facets of the initiative cited in the September report include: l a rewrite of the agency's preapproval inspection compliance program giving the field increased inspection discretion l a pharmaceutical inspectorate focused on prescription drugs and other complex or high-risk operations l a plan to increase the use of product specialists to better link submission reviews and cGMP inspections, and l an internal assessment of the center review program for cGMP warning letters.

ä [EDITOR's NOTE: FDA's effort under the initiative to better define and enhance its overall approach to risk and quality management will be addressed in the next issue of "The Gold Sheet."]

The speed at which the quality initiative is proceeding and its scope bear witness to the agency's level of commitment. At the annual PDA/FDA conference in Washington, D.C., in September, FDA managers underscored the point.

For example, Center for Drug Evaluation and Research (CDER) Office of Compliance Director David Horowitz described the undertaking and the resource commitment as "enormous" reflecting "how seriously we are taking this. There are 16 working groups now set up in addition to the steering committee which meets for lengthy weekly meetings. There are well over 100 FDA staff that are devoting a significant proportion of their time to this initiative."

Horowitz himself is co-chairing the working groups on inspectional work planning and the dispute resolution process, and, he noted, staffers from his office are on "probably every one" of the groups.

ä Addressing the initiative's scope, the compliance official commented that "what we are trying to accomplish here is no less than a revolution" in regulation, manufacturing and product quality.

While the initiative was originally designated to last two years, Horowitz stressed that the timeframe was not meant to "suggest that we will have accomplished all of these lofty goals one year from now and we can go back to business as usual." Instead, during this period "what we aim to do is establish a pathway and to take some important steps forward along that pathway…and to make sure people know that we are serious about this at FDA - that this is not another passing fad and this is not just rhetoric."

Horowitz joined Associate Commissioner Lester Crawford in stressing at the PDA/FDA conference that the initiative extends beyond GMPs to involve "all aspects of FDA's regulation of product quality," including the submission and review of chemistry, manufacturing and controls (CMC) in applications, the inspection process, and setting and reviewing standards applicable to both areas. Actively participating in the initiative are CDER, the centers for biologics (CBER) and veterinary medicine (CVM) and the Office of Regulatory Affairs (ORA) which overseas the field inspection operations.

ä The initiative is having an effect on the pace of guidance development in particular.

All five of the newly released guidances moved relatively quickly through the pipeline, and the agency intends to digest comments and bring the drafts to final form as soon as possible.

Each of the guidances in different ways speaks to FDA's commitment to encourage and facilitate technology and quality science upgrades.

PROCESS ANALYTICAL TECHNOLOGY

The draft guidance on "PAT - A Framework for Innovative Pharmaceutical Manufacturing and Quality Assurance" focuses heavily on "principles" and relatively less on specific recommendations for PAT submissions.

The draft guide's first four sections ("Introduction," "Guidance Development Process and Scope," "Background," and "PAT Framework") delve deeply into the PAT strategy developed by the agency and its rationale for supporting the technologies. The last section ("PAT Regulatory Approach") provides a general overview of how to handle pre- and post-approval PAT submissions and other correspondences.

The lengthy discussion of principles is meant to highlight "technological opportunities and developing regulatory processes that encourage innovation," FDA explains in the guidance. The agency acknowledges that it is "not typical" of its guidance documents to focus so heavily on principles.

The guidance essentially captures and reinforces the podium policy on PAT that has taken shape over the past few years at various industry and agency conferences and advisory committee hearings ("The Gold Sheet" February 2002 and June 2003).

In the "development/scope" section, FDA credits "extensive public discussions" at these conferences as being an "integral part" of the guidance development process. The draft explains that the "discussions covered a wide range of topics including opportunities for improving pharmaceutical manufacturing efficiencies, existing barriers to the introduction of new technology, possible approaches for removing both real and perceived barriers, and many of the principles described in this guidance."

FDA encourages public discourse on the guidance via written comment. On Oct. 21, the agency's pharmaceutical advisory committee will meet to discuss the document.

FDA states that the guidance applies to products regulated by CDER - except those covered by the new Office of Biotechnology Products - and CVM. "All manufacturers of drug substances and drug products (including intermediate and drug product components) over the life cycle of the products" are covered by the document. FDA plans to explore the possibility of expanding the scope of the PAT guidance to biological products.

ä While repeating the disclaimer made at public meetings that PAT is not mandatory, the draft makes clear that it is highly desirable for firms to adopt advanced manufacturing controls, not only from the quality/regulatory perspective but also from a broader public health point of view.

In the development/scope section, FDA states that "any decisions on the part of a manufacturer to work with the agency to develop and implement PAT is a voluntary one" and such decisions do not "mean that similar technologies must be developed and implemented for other products."

However, FDA strongly endorses the technology in the "background" section of the document, asserting that "industry's hesitancy to broadly implement new pharmaceutical manufacturing technologies is undesirable from a public health perspective. The health of our citizens and animals in their care depends on the availability of safe, effective, and affordable medicines. Efficient pharmaceutical manufacturing is a critical part of an effective U.S. health care system."

The agency explains that the technologies afford "significant opportunities" for manufacturers to "improve efficiency of pharmaceutical manufacturing and quality assurance through the innovative application of novel product and process development, process controls, and modern process analytical tools." FDA further states that, "in the future," manufacturers "will need to employ innovation, cutting edge scientific and engineering knowledge, along with the best principles of quality management" in order to handle the various scientific challenges facing the industry. The agency recognizes its own role in helping these technologies come to market and states that the quality initiative is part of that effort.

ä The last portion of the "background" section explains how PAT fits into a risk-based approach to drug quality regulations.

The agency outlines risk-based characteristics of the "desired future state of pharmaceutical manufacturing": l quality and performance ensured via design l specifications based on a "mechanistic understanding" of formulation and process l continuous real-time quality assurance l regulatory processes "to accommodate" current scientific innovations, and l risk-based regulatory approaches recognizing sound science and the ability of "process control strategies to prevent or mitigate the risk" of quality problems.

FDA Defines Framework For PAT

The section on "PAT framework" discusses how the technologies can help firms better "build quality into products."

Subsection A of this section, on "principles and tools," describes in detail the current manufacturing paradigm and then discusses the potential impact on the paradigm of PAT tools, process understanding, risk-based and integrated systems approaches, and real-time release. The discussion "outlines broad principles for addressing anticipated scientific and technical issues" involved with implementing the technologies.

ä Subsection B of the "PAT framework" section addresses FDA's strategy for facilitating implementation.

The PAT initiative requires no changes to the regulations, FDA asserts. Current regulations "can effectively support innovation…as long as clear communication mechanisms exist between the agency and industry." Examples of these mechanisms include meetings and informal communications.

The strategy is four-pronged: the PAT team approach to CMC review and GMP inspection; joint training and certification of review, inspection and compliance staff; scientific and technical support for the PAT team; and the use of guidances to put forth recommendations to industry.

As part of the effort to generate scientific support for the "PAT Review, Inspection and Office of Pharmaceutical Science Team" (PATRIOT), FDA has made cooperative agreements with Pfizer and the National Science Foundation's Center for Pharmaceutical Research to help the agency assess the technology ("The Gold Sheet" June 2003). These agreements were highlighted by FDA in the one-year report to the drug quality initiative, available at www.fda.gov/cder/gmp.

ä The draft guidance reinforces the message that "ideally PAT principles and tools should be introduced during the development phase." The advantage of doing so is to "create opportunities to improve the mechanistic basis for establishing regulatory specifications," FDA maintains.

While various proponents of the technology have espoused PAT implementation during development, the companies working with it indicate that, initially, implementation is likely to occur post-approval ("The Gold Sheet" June 2003).

Various Filing Strategies Discussed

The last section of the draft guidance outlines the various filing mechanisms that can be used for PAT submissions. However, the guidance does not indicate which filing would be required for specific PAT-related manufacturing changes.

ä Instead, FDA reasserts in the section that "close communication between the manufacturer and the agency's PAT review and inspection staff will be a key component" of the regulatory approach.

FDA advises that written correspondence "be identified clearly" as PAT-related so that it can be directed to the PAT team, which has been undergoing a intensive training over the past year ("The Gold Sheet" June 2003). All marketing applications, amendments or supplements, on the other hand, "should be submitted to the appropriate CDER or CVM division in the usual manner," FDA advises.

At the September PDA/FDA meeting, OPS Deputy Director Ajaz Hussain advised that only those firms truly prepared to implement PAT should come forward. "Companies that are ready for innovation," he explained, would not be asking the agency things like: "What should be in a PAT-based batch record?" but would be proposing a format for the batch records and seeking the agency's opinion on that.

FDA is willing to assess "proposals prior to submission or implementation to define the type of data needed to evaluate the proposal and provide a mutually acceptable regulatory pathway," Hussain explained. The goal of this approach, he said, is to provide the "flexibility that is necessary for innovation."

In the draft guidance, FDA addresses a number of specific issues regarding the application of PAT to existing processes. For example, the agency advises firms to "consider the effects of PAT on the current process, in-process controls and specifications" when planning to apply the technologies to an approved product/process. The agency also refers users of the guide to its post-approval changes guidance for NDAs and ANDAs when altering a manufacturing process with PAT.

ä Companies can opt for early or late FDA involvement in their PAT programs.

The draft indicates that the technologies could be implemented "under the facility's quality system" with FDA follow-up via a GMP inspection. Companies also could approach FDA for a "pre-operational review" of the new technology under ORA Field Management Directive No. 135. Another approach involves the filing of a changes-being-effected (CBE) or prior-approval supplement to FDA prior to implementation, and, if necessary, a site visit by investigators. Finally, firms could use comparability protocols. The PAT-trained investigators will be responsible for all inspections conducted under these scenarios.

A bibliography at the end of the guidance refers industry to "useful standards" and literature. The standards listed include those from the American Society for Testing and Materials (ASTM) for various analytical tools and procedures. The Good Automated Manufacturing Practices (GAMP) guide for automated systems and PDA's Technical Report No. 33 on new microbiological testing methods are also referenced. Under the literature reference, FDA advises firms to access the PAT page of the agency's website (www.fda.gov/cder/ops/pat.htm).

ASEPTIC PROCESSING

FDA is also pushing for the use of new technology in its the draft guidance on aseptic processing. Interspersed throughout the document are references to innovations and technologies that are widely thought to offer significant improvements over traditional clean room designs.

ä At a session on the draft guidance at the PDA/FDA conference, CDER compliance official Richard Friedman discussed the document's focus on innovation and how it relates to the drug quality initiative.

In accord with the principles of the initiative, Friedman commented, the draft covers high-risk aspects of aseptic processing and "emphasizes the need for a well-conceived design."

For example, he explained, the deployment of air locks "to provide better facility control" and the optimization of the "flow of materials to the aseptic line and increased automation" can result in significant design improvements. Technologies like isolators, barriers, and blow-fill-seal also are discussed in the draft. While not specifically mentioned in the draft, the compliance official cited the use of robotics as another technology that could improve aseptic manufacturing.

ä The language in the draft guidance strengthened the support for technological upgrades expressed in the aseptic processing concept paper upon which the new guidance is based ("The Gold Sheet" February 2003).

For example, the statement in the concept paper that "a well-designed positive pressure barrier isolator…appears to offer tangible benefits over classical aseptic processing" was changed to remove the qualifier "appears."

The only section in the draft guide that was not in the concept paper is one on "alternative microbiological methods." FDA states that "other suitable microbiological test methods (e.g., rapid test methods) can be considered" for in-process and release testing. The agency encourages the use of tests that "demonstrate increased accuracy, sensitivity, and reproducibility."

In another section discussing media and microbial identification, language was added endorsing the use of rapid genotypic (nucleic acid-based) methods, which are known to be "more accurate and precise than biochemical and phenotypic techniques."

The references to rapid microbiological testing in the draft guidance reflect discussions at pharmaceutical advisory committee meetings following the release of the concept paper in 2002. Committee members, Friedman explained, suggested that the concept paper's silence on the advanced methods might be interpreted to mean "that there is no interest at FDA in seeing alternative modern tests methods, so we included this statement in the guidance."

ä The compliance official reiterated the agency's message that new technologies and a strong commitment to GMPs make an ideal combination for successful aseptic processing.

"Sound GMPs" and the use of well-defined "metrics" are "most beneficial to those firms who include automation and enhanced product protection in their design concepts," he explained. The combination of innovation and GMPs "will lead to more predictable and consistent aseptic processes."

As in the PAT draft guidance, FDA points out in the aseptic document the business-related benefits of using innovative technologies. Friedman stressed this point at the conference: "It is fortunate that many of these improvements [offer] long-term financial benefits in that processes will be more robust and reliable. A well-defined process and better cGMP compliance should have the net effect" of reducing product failures and related investigations. In particular, automation of certain functions "can often mean less personnel are needed for the manufacturing operation," he stressed.

With the guidance meant to foster the use of new technologies, Friedman asserted, it is now "up to drug firms to evaluate where their own operations might be modernized and to consider using today's technology when designing new or modified facilities."

Draft Reflects PQRI Input

Although the draft guidance closely resembles the concept paper in format and content, several meaningful changes were made reflecting the active involvement of the Product Quality Research Institute (PQRI) in the process ("The Gold Sheet" August 2003).

For example, recommendations on airflow were altered in the section on isolators. While the concept paper suggested that unidirectional airflow was the norm for both closed and open systems, the revised recommendation is that a turbulent pattern is "normally acceptable" within closed isolators and that unidirectional flow should be used in open systems.

Language explaining the difference between closed and open isolators has been added to the airflow discussion. Closed systems "employ connections with auxiliary equipment for material transfer," the document says. Open isolators "have openings to the surrounding environment that are carefully engineered to segregate the inner isolator environment from the surrounding room via overpressure." The agency also indicates in this section that closed systems are "generally compact in size and do not house large processing lines."

FDA also dropped language suggesting that HEPA and/or ULPA filters should be used in the isolator air handling system. Now, the agency states generally that the air system "should be capable of maintaining the requisite environmental conditions within the isolator."

ä Another notable change from the concept paper is that "unidirectional" airflow is recognized as acceptable in non-isolator Class 100 environments during operation.

In the concept paper, FDA stated that air in critical areas should be "supplied at the point of use as HEPA filtered laminar flow air…and maintain laminarity during operations." FDA also noted that velocity parameters should be appropriate "to maintain laminarity under dynamic conditions." The draft guide, by contrast, states that laminar flow air should be supplied to critical areas, but that "unidirectional" airflow should be maintained during operations, and that velocity parameters should be appropriate to maintain "unidirectional" airflow during processing.

Friedman pointed out that airflow terminology in the guide is now in harmony with standard industrial usage. "If we don't have the same terminology," he said, "where we are coming from in communication is undermined." The concern was discussed at advisory committee meetings and within the PQRI work group.

A number of significant changes were made to the "air classification" table provided in the "building and facilities" section of the document. These changes harmonize the microbial expectations with the EU's "Annex 1" and incorporate ISO particulate air cleanliness classifications. Friedman commented that the revisions derived from the PQRI report and are "quite an accomplishment."

Another significant change in the table is the replacement of the term "action limits" with the term "action levels" to describe the acceptable environmental quality. Use of the former term by the agency caused much consternation within the industry, which strongly recommended adoption of the latter term. The table also provides both air and settling plate action levels.

ä As part of its involvement in the drafting process, PQRI conducted a survey of industry on current aseptic practices. The survey results had a significant impact, in particular, on the draft's discussion of media fills.

Data on 606 media fill runs during 2002 indicated that only 9% resulted in the finding of contamination; of those, 66% had only one contaminated unit. Based on this data, PQRI proposed, and FDA accepted in the draft, the following three-pronged approach to interpreting media fill results according to the number of units filled:

Fill size	Units contaminated			Action
<5000		0	--
5000-10000		1	investigation/consider repeat media fill
		2	cause for revalidation following invest.
>10000		1	investigation
		2	cause for revalidation following invest.

The concept paper, on the other hand, only specified that "a single contaminated unit in a 10,000 unit media fill batch should be fully investigated, but is normally not considered on its own to be sufficient cause for line revalidation." The data generated by PQRI was an "important cog in the process" of clarifying this section, Friedman commented.

In line with the quality initiative goals of creating a more flexible and risk-based regulatory environment for manufacturers, the use of "qualifiers" and "latitude phrases" was increased in the concept paper and further expanded in the draft guidance. Friedman explained that FDA wanted to "loosen up the language in many places while still stating all the important principles." As a result, the document acknowledges, "there may be prevailing standards that should be the minimum for the great many of the applications, but there are exceptions."

ä The EU has recently changed its guidance on aseptic filling environments provided in Annex 1 of the EU GMPs. The purpose of the revision is to harmonize with ISO 14644-1.

While the table is similar to FDA's, it includes limits for both .5 and 5 micron particles, whereas FDA specifies limits only for particles >.5 microns.

In commenting on the EU revision earlier this year, PDA asserted that the 5 micron limits do not match ISO 14644-1 and are "not practical because the number of 5 micron particles is too small to be statistically significant, and may be confused by noise."

In its written comments (available at www.pda.org), PDA expressed concern with a number of the changes to the EU annex. For example, PDA asserted that "there is confusion about strategic qualification and requalification testing versus operational monitoring." Also, the association took issue with the "mandated use of continuous particle monitoring systems in Grade A and B zones, stating that there is no justification for such a requirement. [continued on p. 12]

FDA Response to Industry Questions On Filling-Line Isolators

While isolators offer significant quality control advantages, like other new technologies they also present challenges in applying and interpreting cGMP expectations. At the 12^th annual ISPE barrier isolator conference in Arlington, Virginia, in early June, FDA compliance officials Richard Friedman (CDER) and Robert Sausville (CBER) responded to a list of 24 questions on filling-line barrier isolators prepared for the conference that delve into issues beyond the specificity of the draft aseptic guidance. Following the conference, Freidman and Sausville edited their responses as appears below. The FDAers addressed a different list of questions on isolators at the ISPE meeting last year ("The Gold Sheet" August 2002).

1. Some companies have actually made the policy decision that they will not pursue barrier isolation technology until the regulatory climate improves. Can you comment on how the FDA is looking at barrier isolation as a mode of aseptic processing?

FDA has been reviewing and improving aspects of the application supplement and GMP inspectional process to provide impetus for more rapid transition to advanced technologies. However, the industry has to take responsibility also and firms need to look internally to see what their own roadblocks might be. Speaking generally, firms are only just recently beginning to submit applications using many of the more modern testing and manufacturing technologies.

2. Is a mechanical integrity test (i.e. a pressure decay test) required on all isolator gloves? Or is a daily visual inspection, combined with a mandatory, periodic glove replacement sufficient to assure integrity? Does the thickness of the glove have any bearing on the replacement frequency or inspection method? Does number of glove interventions per location also have a bearing on the replacement frequency?

Visual checks should be done daily. Our expectation is that mechanical testing should be performed at an appropriate and justified frequency, as visual inspection has obvious limitations. Thickness and other durability attributes are issues that certainly can impact on a science-based schedule for glove replacement. Whether the most frequently or intensely used gloves need to be replaced at accelerated intervals also needs to be assessed. This and other maintenance issues will be addressed during a GMP inspection.

3. Are personnel required to wear sterile latex gloves when an isolator is opened and repair work is being conducted?

Not necessarily. A firm should define and justify appropriate procedures for such repair work when it is done on an isolator prior to cleaning and decontamination. It might be a good idea for firms to use gloves in that a person would not leave skin or oils behind (i.e., in terms of facilitating subsequent cleaning and decontamination).

4. The first air rule in aseptic processing is defined as the requirement for HEPA filtered air to pass over critical component/product areas before the air contacts personnel gloves or equipment obstructions. In some isolators the very act of placing hands into gloves in many glove port locations violates the first air principle. With the encouraging environmental performance of isolators to date, is it a concern of the FDA to break the first air rule in an aseptic processing isolator?

Aseptic technique is still a necessary and vital part of aseptic manipulations in an aseptic processing isolator. Isolator gloves are usually situated close to the equipment because of the "compressed cleanroom" that an isolator essentially represents, but they should not hang directly over the exposed sterile product. We have seen myriad acceptable isolator configurations. These adhere to the CGMP regulations, which require that a firm carefully design procedures to protect the sterile product. Gloves are considered by many with experience in this technology as having the greatest potential for breach, so appropriate measures need to be taken as a prevention strategy.

5. What are the minimum cleaning requirements for change parts before introducing them into the isolator? Does the type of barrier isolator cleaning and sanitization cycle determine the method for cleaning the change parts before introducing those parts into the isolator?

The methods are largely if not totally dependent on the procedures to clean the isolator, and the change parts themselves. We would need more specifics to provide a more direct answer. The principle to keep in mind is that cleaning should be adequate to facilitate subsequent decontamination or sterilization, whichever is applicable. The cleaning validation study data will establish whether the cleaning procedures used are adequate.

6. Are smoke tests required annually for all isolators? If not, is there a stated frequency that this should be performed?

No, smoke testing is expected as part of initial qualification studies. Change control procedures should address the need for further smoke studies in the event that there is a modification in the isolator that would merit new studies. Some firms, as a very good manufacturing practice, do choose to evaluate or verify air patterns of their aseptic processing line on a periodic basis, but there is no requirement to implement this approach.

7. When constructing a barrier isolator facility the airlock access to the filling room (typically a class 100K environment) must be considered. Are there any potential issues with a class 100K space that passes all environmental testing which permits direct personnel access to an unclassified area (i.e. inspection room downstream) without an airlock?

This situation depends on the design and application of the given isolator. We generally have seen an airlock or anteroom leading into the isolator room. Many firms use this area also as a gowning/degowning room. (See #10 for more info on gowning.)

8. Depyrogenation tunnel pressure balance is critical to the maintenance of pressure or vacuum in the filler isolator. Complicating this balance is the fact that most facilities segregate the in-line component washing and filling operations into separate rooms. Is there a negative to using a Class 100K/ISO Class 8 environment for the entire filling operation to reduce facility HVAC zone pressure changes? In other words, for an isolator facility can a washer/tunnel be in the same room as a filling machine?

Designing the facility so that room conditions can be readily maintained and controlled is an important part of CGMP. Separation of functions also would need to be appropriately conceived, per CGMP. A firm should be prepared to justify the design and its robustness given the specific application and facility factors. Note: Recent EU experience from the audience was to separate the wet process (vial washing) from the filling operation. Sometimes washers will leak thereby adding more variables to an already busy process.

9. What is the acceptable air classification surrounding an isolator? If Class 100K, is that defined as 100K during activity (Grade C/ISO Class 8) or 100K at rest (Grade D)?

Class 100K operational is our general expectation, however we would not necessarily rule out the possibility of a well- designed and controlled isolator housed in a Grade D cleanroom.

10. What level of gowning would the FDA expect of personnel working around a continuous filling open isolator in a class 100K room (Grade C/ISO Class 8)?

FDA has not prescribed specifics on the issue, preferring instead to address this question in more general terms. For example, the concept paper on aseptic processing simply states: While cleanroom apparel requirements are generally reduced, the contribution of human factor to contamination should not be overlooked.

11. It is an accepted practice to locate a non-viable particle monitoring location in close proximity to an exit mousehole on an open isolator. Is there any requirement to also provide a viable sampling location (settle plate or active microbial sampling) near the mousehole? This would be in addition to the microbial sampling location placed within the "critical zone" of filling.

Our focus would be on a firm's overall risk-based environmental monitoring approach, rather than specifying any single location that we would expect to be monitored across the board for all isolators. But we agree that this location is a very valuable critical control point to monitor.

12. Viable monitoring practices in barrier isolators vary, with some using settle plates, most using SMA-type active air monitoring and surface swabs or RODAC plates. Does the FDA agree that introduction of such growth media into the isolator system during production presents a risk to the product exposed on the line at the time? Has industry data convinced FDA that reduction of such testing or limiting testing to the end of a campaign are acceptable measures?

Firms with isolators normally have a lessened EM burden because many of the surface samples that have been taken daily for traditional lines can only be taken at the conclusion of a campaign for an isolator. While many samples are taken at the end of a campaign rather than with each batch, there are nonetheless sound methodologies available to conduct other tests, such as air monitoring, at appropriate intervals throughout a campaign. An appropriate balance should be affected in conceiving the EM program. As a general principle, you want as much information as possible for QA/QC and production to evaluate a deviation or failure when one occurs . When things are operating well you know it because you have data. And when a problem does occur, you will want to know its extent -- when did the problem start?

13. For a depyrogenation tunnel connected to a barrier isolator, is environmental microbial testing required in the tunnel cool zone? If so, how often should it be tested - every campaign? once per year?

A firm should make a risk-based decision on EM locations. As we've stated in the past, the cooling zone should not be a source of contamination, and if there is a concern that it may convey contaminants, then this might be a location to assess at a frequency appropriately defined and justified by the firm.

14. Surveys have indicated that half of the people using barrier isolators have sterilizable cooling zones in the depyrogenation tunnel. If one does not have this sterilizable feature, what is the required classification surrounding the tunnel? Is a different classification requirement if you don't have a sterilizable cool zone?

It is one of the many considerations a firm should incorporate in their decision on what background environment is appropriate for housing their isolator. Additionally, as stated earlier, the depyrogenation tunnel should not be a source of contamination. We have seen microbial contamination problems attributed to a cooling zone that was not sterilized or sanitized after mechanical work was done. A firm needs to have appropriate procedures in place to prevent this route of contamination.

15. What are the FDA's expectations in terms of D-values for biological indicators used to demonstrate decontamination efficacy of VHPÔ and other gaseous sterilants? Are there any D-value methodologies that they would prefer to see used? Must the D-values be performed on all internal isolator surfaces, e.g., stainless steel, Plexiglas, Tygon, etc.?

The Sigwarth/Stark paper presented in the conference, and published in the PDA Journal, is very useful. At the design stage, it is advisable to start by reviewing such published work to help in choosing materials of construction that are most amenable to an efficient and robust decontamination. A firm should select sufficient representative material surfaces to evaluate during qualification/validation. A D-value methodology has not yet been standardized, but we have seen firms use well-controlled and closely monitored small isolators with controlled levels of H₂0₂ (at a specific concentration) to help develop the microbial lethality curve used to estimate the BI's D-value. An important factor in performing these determinations is proper preparation of your BIs.

16. Barrier isolators typically consist of many non-product contact components (both filling machine and isolator structures) covering many materials of construction. Can D-value studies be reduced by performing cycles based on the texture and porosity of a surface rather than material of construction?

Texture and porosity are generally quite influential, but the latest research data also indicates that material composition can also be of high significance. For example, hypalon is a rather rough surface but organisms are killed relatively easily; whereas POM is smoother but it is more difficult to obtain kill of the BI on this substrate. Be careful not to oversimplify on this matter when considering materials because research data indicates that there is a combination of factors at play.

17. Some barrier isolator manufacturers and users have succeeded in decreasing sanitization cycle times by directing hydrogen peroxide gas into the critical zone, bypassing the recirculation HEPA filters. This reduces the absorption of sterilant (i.e. VHPÔ ) onto the filter and decreases the aeration time required to achieve the target peroxide level for subsequent aseptic processing. Is it a requirement that barrier isolator sanitization cycles actually pass sterilant through the HEPA filters?

This practice would be acceptable as long as all interior surfaces are decontaminated. If the process is not robust, and there is inadequate distribution of the decontaminating agent, then that would raise questions.

18. What is the expectation of the agency with regards to performing a "set-up" media challenge for initial contamination challenge of an isolator? Is it required to perform both "set-up" media challenges and "piggyback" media challenges (a challenge at the end of the isolator campaign duration) each year on a given isolator filling machine?

The overall design of the media fill program should take both setup and process length aspects into account. Setup should be addressed at a suitable frequency but does not necessarily have to be addressed in each media fill. The overall aseptic requalification program should ultimately provide assessment of control points and variables of the process and determine the state of control (i.e., qualification) of the line.

19. What is the maximum campaign length with which the FDA is comfortable for a barrier isolator system? (i.e.14 days? 21 days)? What are the acceptable media challenge unit quantities and acceptance criteria to warrant this relatively prolonged isolator filling campaign duration? What is the current thinking on the maximum duration of use for liquid product sterilizing filters?

We do not prescribe any specific maximum campaign length. We have often seen week-long campaigns. Whatever the campaign length selected, we will ask to see data. As campaign length increases, it is more challenging to address all of the relevant issues in a media fill. Tough questions remain on the maximum duration of use for sterilizing filters. These filters need to be changed at appropriate intervals, and a firm should safeguard against the possibility of grow-through.

20. Has the FDA developed a training document or program for inspectors that are involved in isolator inspections? Is that document available to industry?

Isolators have been discussed at drug and biologics training courses, but the impending publication of aseptic processing guidance will go a long way toward providing a common ground for industry and FDA in addressing isolator CGMP issues.

21. Due to the recent significant turnover in the FDA, the need is noted within the sterile processing industry for understanding and training of inspectors within the FDA. This is especially true with the changes in technology to which the industry is current moving. How is the FDA going to train and retrain inspectors on emerging issues?

This is best answered by ORA, but we are aware that there are new and innovative approaches being used, including computer-aided training. Once we have published the Aseptic Processing Guidance, we plan to provide training for the Field on the new guidance. We are also plan to update our sterile drug inspection compliance program once the aseptic guidance is published.

How does the FDA monitor inspector's capabilities? Again, a topic that can be best addressed by ORA. Investigators have to meet certain training requirements, testing has been implemented to confirm comprehension during our training courses, and further certification procedures are being developed at this time for Drug and Biologics in accord with the CGMPs in the 21^st Century Initiative. Perhaps most importantly, we are in the process of creating a specialized cadre of expert investigators who will be members of a new Drug Inspectorate.

What methods are being used for open/honest review of inspections between industry and FDA? We have always been available through informal measures for interaction on 483 issues with industry, however we recognize that more formal procedures may serve our common goals. We are currently working on a formal dispute resolution process that will further address concerns regarding unresolved technical matters.

What is being done for consistency between US inspections and overseas inspections? Different inspectorates all over the world use varying approaches at times, but have more commonalties than differences. For example, CGMPs regulations are quite similar in most regions and inspections all have the same goal - to observe conditions of the facility and determine the ability for a firm to consistently and predictably produce a product meeting its pre-determined attributes. We are not prepared to comment further today on this issue as it is most appropriate for international staff in the Centers, ORA, and Office of Commissioner to address these matters more exhaustively.

22. Could the FDA share some examples of problems that they have identified with isolators, either here or abroad, e.g., poor design, retrofit of an existing line, microbial contamination problems, validation problems, etc.? The problems seen could be changed a bit to illustrate the point, but not reveal the source [firm] where the problem occurred. Sharing some common problems seen in isolators with industry might help prevent them from being repeated at other firms.

There is a need for establishing good measurements for detecting changes in the isolator's dynamic environment. Some examples include pressure differentials, particle counting, as well as microbial data.

Firms are generally doing a very good job of controlling and maintaining gloves through mechanical integrity tests and well-conceived preventative maintenance programs. However, one firm had gloves that were contacting sharp edges. The gloves wound up repeatedly being punctured and the firm ultimately improved the equipment design to address this cause. At another firm, our PAI inspection found that glove breaches had led to multiple media fill failures upon initial aseptic line qualification. Of course this is only a cautionary example and is by far the exception to the rule with isolators, which have established a tremendous track record as mentioned in my presentation. The problem in the particular case was this firm thought going into these initial media fills that since this was an isolator, small holes in gloves were not going to be a problem. The firm responded by improving their glove control procedures, ran successful media fills (no contamination), and received approval very shortly thereafter. So it is just important that proper CGMP procedures be written by those who are very familiar with the isolator, its equipment, and its operation.

Remember risk-based approaches often work very well when developing isolator procedures. Also, if obscured, surfaces may not be effectively sanitized during decontamination. A circumstance [occurred] wherein during decontamination of a sterility test isolator, a glove had contact with the floor of the isolator. The area later tested positive for Bacillus sp..

23. What is the expected response when a pinhole is detected in a glove during product filling in a Restricted Access Barrier System (RABS) with grade B background, assuming that the system has been successfully challenged with pinholes in gloves?

An investigation. The response varies with the situation. The firm should look at origin and any product impact, etc.. Some other considerations for the investigation: Where was the glove used, what was it used for, how frequently was the glove used? Does a leak potentially constitute a breach in this particular RABS system or significantly undermine product protection?

24. During cleaning and equipment set-up, is it acceptable to open RABS doors into the grade B background, or is a grade A 'border' required to protect the internal environment when a door is open?

It is hard to answer a question where the definition of the term has not been fully elucidated. The difficulty here is that there is not a precise definition of a RABS. There are basic barrier concepts in use throughout the industry right now for nearly every aseptic filling line. They run the spectrum from limited barriers to extensive barriers. The more extensive barrier concepts that have been loosely termed as RABS seem to allow for a lot of variations in meeting the intended purpose of an aseptic operation - i.e., protecting the product. It would be useful to better define the attributes and standards that are at the essence of any RABS system. Perhaps even different types of RABS can be defined, such as a Type 1 RABS and Type 2 RABS. But we think that it will be important to articulate objective standards that define such systems. ISPE may want to serve the important role of driving efforts for a clearer definition so our discussions begin with a common understanding of terminology.

We are not concerned about doors opening during cleaning, and prior to disinfection or decontamination of such enclosures. But, once aseptic setup or aseptic production activities are in progress it would seem that a basic threshold for claiming of a RABS unit would be that there is at least class 100 cover over doors and gloves at all times. We are assuming that the model being discussed in this specific question is one in which doors swing open into the room environment and gloves are installed on these doors. If the gloves were to swing into a Class 10,000 environment, it would seem to be much more of a classical or traditional aseptic operation. So one would think there would generally be Class 100 HEPA filter cover over the relevant area.

PART 11

FDA's revised policy on electronic records/signatures (21 CFR Part 11) fits in with the emphasis placed on new technologies by the drug quality initiative.

In February, FDA published a draft guidance on the "scope and application" of Part 11 that spelled out which provisions of the rule would and would not automatically be enforced.

ä The final guidance clarifies many of the questions raised by industry in commenting on the February draft.

At the PDA conference, CDER compliance official Joseph Famulare discussed the public comments and the revisions they prompted. He pointed out that 61 commentators submitted over 300 written comments.

One of the clarifications made in the final version of the guidance is that Part 11 and computer validation requirements "still remain in effect," Famulare stated. "We clearly delineated all the sections where enforcement discretion is not being applied, in addition to saying all areas were it is being applied."

The final document also better defines "enforcement discretion." Famulare explained that "there were some questions about what do you mean by enforcement discretion. We got a little bit more specific in our wording in the final guidance that we do not intend to take regulatory actions on these issues."

In his remarks during the plenary session of the PDA/FDA conference, Horowitz explained further the concept of "enforcement discretion" as applied to the electronic records rule: "We mean we do not intend to devote enforcement resources to certain key parts of Part 11 that we felt were being misunderstood and misapplied; in particular, validation, audit trails, copies of records, record retention. And we tried to provide more clarity that we don't intend to enforce any of Part 11 for qualifying legacy systems."

ä In response to industry concern, FDA removed a reference to a NIST document on risk management for IT and instead refers in the final version to an ISO risk management guide for medical devices (14971).

Famulare noted that the NIST document "was more based on IT system applications" whereas the ISO document is "a little more appropriate for what the guidance was covering." Horowitz asserted that the ISO document is a "helpful tool in thinking about risk" and that it "will be very helpful in applying quality control and quality assurance to electronic records."

Commentators also raised concern about the previously issued Part 11 draft guidances which were withdrawn. Famulare indicated that there are no current plans to reissue those documents.

The final document clarifies the meaning of "legacy system," Famulare pointed out. "We go through some criteria as to what constitutes a legacy system and what happens when there are some changes to that system." The final version repeats language included in the draft that a "legacy system still has the overriding need to meet the predicate rule requirements, be fit for its intended use, and if it has changed you need to evaluate those changes in terms of meeting the predicate rule requirements," he stressed.

Famulare emphasized that, as a general scope and applications guidance, it "does not answer every possible technical type question that one can come up with Part 11 systems and applications."

FDA will reexamine the rule itself, an exercise that will "probably lead to a rulemaking process" to revise the regulation, Famulare maintained. The agency is committed to "stay on this course" and open the rule revision process to the public, he said.

The hope, the compliance official indicated, is that by using the guidance there will not be "hesitation by companies to innovate and make rationale decisions about how Part 11 is applied for computer systems."

COMPARABILITY PROTOCOLS

FDA views the use of comparability protocols as an important tool in facilitating manufacturing upgrades, particularly for therapeutic biologicals.

The September draft comparability protocol (CP) guidance for proteins and other biologicals is the companion to the draft CP guidance for drugs (including well-characterized synthetic peptides) released by the agency in February ("The Gold Sheet" April 2003).

The new draft does not pertain to blood and blood components nor vaccines for veterinary use. A discussion of comparability protocol use for blood products can be found in FDA's July 2001 guidance on "changes to an approved application" covering these products.

ä The two draft CP guidances are nearly identical in structure and content. There are a few relatively minor editorial changes to the later guide and some material added specific to protein/biological products.

The public comment period on the drug CP draft guidance closed in June. Comments on the bio version are due by December 4. FDA plans to finalize the drug document in early 2004.

The bio draft includes a definition of a comparability protocol not provided in the drug version. A CP, FDA clarifies, is "a comprehensive plan that describes the specific tests and validation studies and acceptable limits to be achieved to demonstrate the lack of adverse effect for specified types of manufacturing changes on the identify, strength, quality, purity, or potency of the product, as they may relate to the safety or effectiveness of the product."

The new draft adds that FDA's review of the CP "will include a determination of whether changes made in accordance with that protocol may be submitted under a reduced reporting category for the change because the use of the protocol reduces the potential risk of an adverse effect."

Both draft guides discuss specific concerns that should be addressed in CPs covering changes in analytical procedures. The bio version adds the caution that when the revised analytical procedure is being used for release or process control, "you should not delete a test or relax acceptance criteria that we approved in your application, unless and until FDA informs you that the approved acceptance criteria are no longer required."

ä The CP information specific to biologics incorporated into the document includes: l additional references to supporting FDA guidances l the potential impact of the added facility/equipment information in biological license applications (BLAs), and l various examples of previous protocol use for biologic products (see box below).

Examples Of CP Use For Biologics

The September draft guidance provided the following list of specific examples of biological process changes for which FDA has received comparability protocols.

• Increase or decrease in batch size that affects equipment size,

• Modification of production operating parameters in fermentation (e.g., time, temperature, pH, dO₂ (dissolved oxygen)),

• Adding, deleting, or substituting raw materials (e.g., buffer or media components),

• Mode changes (usually associated with equipment changes such as tangential flow filtration to centrifugation),

• Establishing a new working cell bank using a modified procedure,

• Reprocessing the drug substance or drug product, as appropriate,

• Addition, deletion, or rearrangement of production steps; and

• Facility-related changes for products with facility/establishment information provided in a BLA, or postapproval supplement to a BLA (see examples provided in Section V. E.).

FDA notes that information on assessing the effect of CMC changes, supporting documentation needed and reporting categories is provided in its guidances on "changes to an approved application" for biological products (1996) and specified biotech products (1997) as well as in its 1996 guidance on demonstrating comparability of biologic/biotech products. The reference to the Scale Up and Post-Approval Changes (SUPAC) guidances is removed from the bio version as they apply only to chemically derived drugs.

The agency explains that "for biologics, which also have application requirements described in an Establishment Description section, there may be additional situations when a comparability protocol can be useful." In particular, the draft notes that the CP has been beneficial when introducing additional protein/ biologic products into an approved dedicated area in a facility.

FDA cautions, on the other hand, that the agency's ability to assign a reduced reporting category may be limited when the facility information is provided in the BLA. The draft cites as examples major changes in equipment or utilities, and the introduction of additional product(s) into an approved dedicated manufacturing area of a facility where containment is a concern.

CPs Apply To Single, Multiple Or Repetitive Changes

As in the February drug CP guidance, the bio draft explains that a comparability protocol can describe a single CMC change or multiple related changes, and "can be particularly useful for changes of a repetitive nature."

In comments submitted to FDA in June on the drug version, PDA suggested that FDA should clarify the ability of firms to "bundle" the same or related changes across multiple products/applications. The association noted that the bundling approach is now "well-established;" for example, in changing solid oral products to a new packaging system.

PDA also suggests that FDA's CP guidance should clarify how a comparability protocol is used when a drug master file (DMF) is involved.

In general, PDA commented, comparability protocols "would be more useful to manufacturers if FDA could provide data requirements for some common changes." Examples of such changes cited by the association include an alternate API supplier, an API manufacturing site change, an alternate testing lab, a product line extension such as a new fill size, an expiration dating reduction and a stopper change. PDA provided three examples of such potential data requirements in an attachment to its comments.

ä How the comparability protocol approach might dovetail with the agency's "bulk active post approval change" (BACPAC) initiative, in particular, is receiving attention now by a "specifications" working group serving under the Drug Substance Technical Committee of PQRI.

The PQRI working group was created specifically to develop recommendations on the data and filing type provisions that should be included in the agency's forthcoming "BACPAC II" guidance. BACPAC II will address drug substance manufacturing changes after the isolation of the final intermediate. A draft of BACPAC I (changes before the final intermediate) was published by FDA in late 1998 ("The Gold Sheet" December 1998 and May 1999).

Consideration of comparability protocols in drafting BACPAC II was recommended by participants in a specifications working group-sponsored breakout session at a PQRI public workshop held in early August ("The Gold Sheet" August 2003). Based on those discussions, the PQRI working group decided to examine the issue in coordination with BACPAC working groups operating under the Pharmaceutical Research and Manufacturers of America (PhRMA) and the Generic Pharmaceutical Association (GPhA).

Application CPs Require In-Depth Product Knowledge

The new draft stresses that, because of the complex and heterogenous nature of protein/biologic products, knowledge of how product attributes affect safety and efficacy is "crucial" in CP design.

The bio version further cautions that the appropriateness of submitting a CP with the original application needs to be carefully evaluated "when your experience manufacturing the product is limited and it may be difficult to identify" the appropriate CP elements.

ä In its comments on the drug CP draft, PDA cited the inclusion of planned changes in initial NDA/ BLA submissions as an area warranting more explicit FDA advice.

The association suggested that the guide could provide information, for example, on the impact on the review cycle and the location of the information in the Common Technical Document.

The issue is an important one, PDA explains, as companies often need to optimize manufacturing processes soon after approval of the NDA or BLA. The association pointed out that these changes to the manufacturing process for a drug product or active pharmaceutical ingredient "serve a variety of useful purposes, such as quality improvement, waste reduction, efficiency enhancement, etc. The ability to reasonably predict the process will significantly improve implementation by providing a predictable timeline for successful implementation."

The need for a more accommodating regulatory approach to the manufacturing refinements firms want to make as they gain full-scale experience was a focal point at the April conference on the drug quality initiative cosponsored by FDA and PQRI. Along with the use of comparability protocols, the conference participants considered a proposal for the submission of a one-time "super-supplement" to report all changes made to a process during a product's first year of marketing ("The Gold Sheet" May 2003). The super-supplement concept could be linked to an application CP approach through which the changes would be previewed, or could preclude the need for such CPs altogether.

While greater use of comparability protocols was among the application-based solutions to the manufacturing change problem proposed at the FDA/PQRI conference, participants pointed to the limitations of this individual application, reviewer-oriented approach as an efficient mechanism for broad industry regulatory relief.

At the September PDA/FDA conference , CDER's Horowitz noted that the comparability protocol approach lines up with "many of the themes" of the overall GMP/quality initiative including: allowance for manufacturing changes without submitting prior approval supplements; facilitating continuous improvement and innovation; and targeting regulatory review according to the risk that a manufacturing change will adversely affect product quality.

The compliance official added, however, that the agency is "actively considering other options for managing manufacturing changes without the need for prior FDA review."

DISPUTE RESOLUTION

The draft guidance on "Formal Dispute Resolution: Scientific and Technical Issues Related to Pharmaceutical cGMP" better defines the multi-tiered plan proposed by the working group in February ("The Gold Sheet" April 2003).

Like the initial proposal, the draft guide provides for an "informal" dispute resolution involving the firm and the visiting investigator and a two-tiered "formal" process extending beyond the investigator level.

As previously specified, the first tier of "formal" resolution involves the district offices and the relevant centers. The second tier involves an "agency level" dispute resolution panel comprised of "neutral experts" to whom firms can appeal decisions reached during the tier one process.

ä Unlike the initial proposal, the draft offers specific timeframes for pursuing the formal aspect of the dispute resolution plan.

Should informal negotiation with the investigator during the GMP or preapproval inspection fail to satisfy the company, it would have 10 days to initiate tier one of the formal process. The district offices and the centers would, in turn, have 30 days to respond with either a decision or a reason for a delay in resolving the issue. If the tier-one decision is unsatisfactory to the firm, it would have 60 days from receipt of the decision to appeal to the panel.

A section of the draft guidance explains how to make requests for dispute resolution, including addresses of relevant offices. Another section discusses what information to include in requests and how the requests should be identified.

The dispute resolution group noted in its six-month report that it was looking into the feasibility of including into the plan a 48-hour delay in the issuance of inspection reports if a dispute occurred. The draft guidance, however, makes no such allowance, indicating that such a delay does not conform with the regulations.

ä FDA is going to conduct a pilot of the technical/scientific dispute resolution plan as outlined in the guidance. The pilot is meant to help the agency and industry evaluate the plan with hands-on experience in order to make more informed comments/revisions to the draft.

CDER's Horowitz noted at the PDA/FDA conference that the pilot will help the agency "evaluate and figure out how to tinker with, adjust and finalize this guidance." Although unable to specify which districts would be involved in the pilot, he stated that it will be "broad" and will start "no later than January 1, 2004."

By providing industry the opportunity to comment on the guidance "as it is actually being used in a pilot," the agency anticipates that the comments "will be more productive," Horowitz explained.

INSPECTION PROGRAM

The "pharmaceutical inspectorate" is taking shape. FDA announced in its one-year report on the quality initiative that it is planning to train a total of 50 investigators for the program by the end of FY 2007.

The program will result in a cadre of trained inspectors specializing in prescription drug manufacturing facilities and will be similar to the development of the specially trained PAT inspection team. The first 10 members of the pharmaceutical inspectorate are to be trained by the end of 2003, with an additional 15 available by the end of FY 2004.

ä A "memorandum of understanding" (MOU) between the Office of Regulatory Affairs (ORA) and CDER outlines the plan for selecting and training the inspectorate.

The program is voluntary and open to qualified candidates, the MOU states. To qualify, volunteers must possess "the necessary critical thinking and communication skills," have "at least three years experience in inspecting pharmaceutical manufacturing," and also be certified as "Level II Drug Investigators."

The district offices will then nominate those volunteers it deems suitable for the inspectorate. After selection, investigators will be trained and certified as "Level III" drug investigators. The criteria for Level II and III certifications are still under development.

Training at Level III will be overseen by a "Level III Drug Investigator Certification Board." The MOU notes that this body will consist of "two field investigators operating at Level III in the area of certification being reviewed, two experts from [CDER], two experts from [CVM], on Division of Field Investigations expert (with appropriate technical expertise and investigation experience, one ORA manager from the appropriate field committee," and one representative from ORA's Division of Human Resource Development.

ä Funding for training and travel will be a challenge for the agency.

Under the MOU, CDER commits to "provide funding (amounts to be negotiated annually with [ORA]) for the development and implementation of training programs identified in the Level III Drug Certification program for the PI," and for "continuing education for the members o the PI to maintain their certification."

ORA will "provide funding to develop the training for the Level II Drug Certification" and form the Level II and Level III Drug Investigator Certification Boards based on membership inclusive of the Center."

The pharmaceutical inspectorate will operate both domestically and internationally, although the number of inspections conducted by individual inspectors outside their home districts is not yet determine. FDA has indicated that the inspectorate will spend 80% of their time conducting drug quality inspections and related activities.

Preapproval Inspection Discretion Provided

FDA also announced in the one-year report that it has modified the chapter of its "Compliance Program Manual" (CPM) on new drug evaluation (chapter 46) to allow the field increased flexibility in deciding when to conduct preapproval inspections.

The changed manual removes the top 200 prescribed drugs and narrow therapeutic index drugs from the list of situations requiring PAIs.

In explaining the policy change at the PDA/FDA conference, Horowitz noted that "there are many cases where we have adequate information from recent, comprehensive, systematic GMP inspections to know enough about the GMP state and the quality control and quality systems in place to dispense with a preapproval inspection."

By giving the field more discretion to use this information, Horowitz said, the agency can "take advantage of that knowledge" and preserve limited resources. He intimated that the CPM revision was "the first step" and industry "will be hearing more about this."

ä FDA also will be reviewing the new procedures for CDER and CVM review of cGMP warning letter review begun in March.

Horowitz announced that the review "already has" facilitated field/center communication. FDA will begin an "internal assessment of the program "shortly," he explained, to identify "the areas where the center had a different approach than the field…and try to reconcile those different approaches with internal directives, more training, and guidance documents."

Update by CDER's Horowitz On FDA's "21^st Century" Quality Initiative

At the September PDA/FDA conference, CDER Compliance Office Director David Horowitz discussed the progress the agency is making in its "21^st Century" quality initiative. After explaining the "challenges and opportunities" that motivated the initiative and its scope, he gave a brief update on the various facets including: l Part 11 l dispute resolution l aseptic processing l pre-approval inspections l comparability protocols l process analytical technology l manufacturing science l risk-based approach l pharmaceutical inspectorate l procedures for cGMP warning letters l international cooperation, and l quality management systems. Horowitz's remarks at the conference complement a "progress report and implementation plan" released by the agency on Sept 3 at the initiative's one year mark.

I am going to talk about what is known as FDA's "GMP initiative" informally. For those of you have been following the issue, I think it is much broader than just GMPs. I am going to talk a little about why the FDA began this initiative, which is about a year old now. And then I am going to talk a litt le bit just about the scope of the initiative. And then I am going to talk about the September 3 progress report which some of you may not have seen. There is a great deal of information that we released on the third of September on our website. There is a lot of information on that website. I am going to go through some of the highlights and then give you the opportunity to ask questions after that.

Why did FDA undertake this initiative? The last major changes to FDA's system for regulating drug quality really occurred almost a quarter century ago when FDA made substantial revisions to the GMP regulations. Since that time there have been some incremental shifts to our approach to GMPs and also our approach to manufacturing changes and application submission review in the FDAMA legislation and with SUPAC. We won't go into that today. But since that time, you are all aware that dramatic changes have occurred in the environment for pharmaceutical regulation. I am going to divide up some of those changes into challenges and opportunities that motivated the initiative.

The first challenge that is relatively apparent to me from a compliance role is the very significant decline in inspectional resources available to FDA. From 1980 our resources available for GMP inspections have declined by about two-thirds. Now this chart doesn't include resources devoted to preapproval inspections which have increased since PDUFA, but I have seen the chart with those numbers as well and there is still a very steady and significant decline. The preapproval inspections in general are not a substitute for a comprehensive systems-based GMP inspection. So we really have to think about whether the system that we had in 1980 works as well with one-third the resources for inspections that we have now available.

Clearly the environment has also changed, and one of the other challenges is the larger roles that pharmaceuticals have played in healthcare. That includes the increased number of products, but it also includes a wider range of drugs being manufactured and using a broader range of manufacturing techniques and for a broader set of purposes.

The globalization of the pharmaceutical industry has also posed significant challenges for us over the last 25 years, one of which concerns the increase in the number of foreign manufacturing sites. You just can't keep up with the increase in globalization that has occurred over the last 25 years. I have heard that at least 70% of the active pharmaceutical ingredients are manufactured abroad.

It become increasingly clear to us in looking at these challenges that the old model needed some revision. Other challenges included dramatic advances in the pharmaceutical sciences, including the application of biotechnology to drug discovery and manufacturing. You might not think of that as a challenge. But from FDA's perspective, we have to be able to regulate a more complex and diverse world of manufacturing, and an increasing number of manufacturing supplements came in during that time period.

But many of these same challenges as is often the case were also opportunities for us, and we tried to look at them as such. And that further encouraged us to think about making major changes in the way we regulate pharmaceutical quality.

The first opportunity is the major advances in manufacturing science and technology. One thing that we noticed was that many of these revolutionary changes were adopted in other industries to a greater extent then they had been adopted in the pharmaceutical industry. The chemical manufacturing industry, for example, was much quicker to adopt some of these changes. That I think created an opportunity for us to think about what we might do - how we might change our approach - to encourage and enhance the adoption of modern manufacturing science and technologies.

There have also been significant changes in what is known to some as the science of quality management. That includes various management techniques, it includes six sigma management tools, and it also includes I think a variety of quality systems approaches which have become increasingly prevelant. To some degree, FDA has begun to encorporate some of these concepts in our expectations of industry as well as our own internal processes. But I think we realized that on both sides we have some opportunities for further progress. HACCP is one systems-based approach, risk-based approach, explicitly incorporated in FDA's device regulations, and it is also explicitly the centerpiece of FDA and USDA's regulation of certain food products, seafood and others. And we in the drug side in the last couple of years implemented a systems-based approach to drug inspection. I would encourage the pharmaceutical industry to think more in those terms about controlling risks and quality by design.

Other opportunities I think include advances in the application of risk analysis and risk management. Over the course of 25 years, the information technology revolution has provided us with tools to analyze data and to take risk management to a completely different level in terms of a systematic and rigorous approach as compared with 25 years ago. Again, I think risk management is used and has been used for some time in the pharmaceutical industry. In fact risk management is something that is part of our daily lives. Just to travel here, you had to incorporate many risk management concepts and risk control and risk assessment concepts. But what we are talking about is doing this in a more systematic and rigorous fashion. And we believe there is an opportunity - an opportunity for both industry to do that and for FDA to do that as well - to better prioritize and focus our work. In fact other regulatory agencies have made substantial strides forward in this area. Similar regulatory agencies like EPA, customs and OSHA have made great efforts to develop complicated models to identify risk factors to target and focus their work on what matters most. Unfortunately, I have to report, IRS has done the same thing.

The scope of the initiative: As most of you probably know, this is intended to be a two-year initiative. It was announced about a year ago, and we issued a six-month progress report in February 2003 and our anniversary report on September 3^rd. I will be talking more about that. When I say this is a two-year initiative, I don't want to suggest that we will have accomplished all of these lofty goals one year from now and we can go back to business as usual. I think what we are trying to accomplish here is no less than a revolution in the regulation and manufacturing and product quality. I think what we aim to do is establish a pathway and to take some important steps forward along that pathway during this two year period, and to make sure people know that we are serious about this at FDA - that this is not another passing fad and this is not just rhetoric.

I also want to make clear, as I alluded to earlier, that this is not just a GMP initiative. This involves all aspects of FDA's regulation of product quality. And that includes submission review, chemistry, manufacturing and control of our approach to regulation. It includes the inspectional programs. And it includes standards setting at both of those stages. It includes setting specifications at the application review stage, and it includes interpreting and applying GMPs through guidance documents and regulations.

It is not just limited to CDER. It directly involves active participation from all of our pharmaceutical centers: CDER, CBER and CVM, and especially active participation from ORA - the operational unit, the eyes and ears of the agency that implements much of the GMP inspectional program. It applies to human and veterinary pharmaceuticals and human biological drugs.

This chart [indicates that] this has become a fairly enormous undertaking for the agency. Enormous resources are being devoted in the agency, just to give you an idea of how seriously we are taking this. There are 16 working groups now set up in addition to the steering committee which meets for lengthy weekly meetings. There are well over 100 FDA staff that are devoting a significant proportion of their time to this initiative. I am co-chair of two of these work groups, and I have got staff who are members of probably every one of them. I focus on the inspectional work planning and the dispute resolution process most directly, but I am familiar with the activities of many of the other groups, and I will talk about some of that.

As I mentioned, the second progress report came out on September 3. Some of the highlights include: l five new guidance documents, four of which are draft, one final, intended largely to enhance the consistency and coordination of the drug quality programs. They accomplish other things as well, which I will talk about. l Also we announced collaborative arrangements with academia, industry and other government organizations which I think are going to be essential for moving this initiative forward. l We announced steps to streamline and improve FDA's internal processes, which we believe is essential for taking our regulation of product quality to the next level. l We announced efforts to improve international standards with enhanced harmonization and collaboration with international counterparts, and I will speak a little more about that as well.

Part 11

First let's talk about Part 11. Part 11 concerns the regulation of electronic records and signatures. It has been very controversial and I think in some respects very misunderstood. We had a great opportunity I think to clarify the scope and our intentions for this regulation. In February we issued a draft guidance. We got a number of very helpful comments, and we issued a final guidance to try to respond to those comments and bring this stage to closure. We announced the intention to initiate rulemaking, however, which is the next step - to amend the regulation itself - continuing on the path that we enunciated in the guidance.

I think most importantly that the step forward that we took encourages the use of electronic recordkeeping and submissions by addressing the uncertainty about regulatory expectations in this area. We have discovered that that lack of clarity and some misunderstandings about the regulation and how it may have deviated from its initial intent had actually discouraged the use of modern electronic recordkeeping in a variety of manufacturing and submissions contexts where it is clearly extremely helpful for efficiency and product quality.

More specifically about Part 11: I think the most important thing is that it narrows the scope of Part 11 by more narrowly defining electronic records that are subject to the rule. It focuses in particular on records that are required to be kept by what are known as predicate rules - that means other regulations like GMPs. So if you see an opportunity to improve your processes and use electronic recordkeeping, that won't necessarily be a Part 11 record. So you are encouraged to do that. It won't be a Part 11 record if it is not required and you do not intend to use it as a record that you will be submitted to the agency.

Also important here I think, the guidance makes clear that FDA does not intend to enforce - now the term enforcement discretion is used widely in the agency, but we want to be clear what we mean by enforcement discretion. We mean we do not intend to devote enforcement resources to certain key parts of Part 11 that we felt were being misunderstood and misapplied; in particular, validation, audit trails, copies of records, record retention. And we tried to provide more clarity that we don't intend to enforce any of Part 11 for qualifying legacy systems….

This distinguishes a little bit what was accomplished in the draft guidance vs. the final guidance. One thing that we did that I haven't mentioned already is that we used the ISO guidance as a reference….The ISO guidance is one of several helpful tools in thinking about risk. And I think that will be very helpful in applying quality control and quality assurance to electronic records and signatures.

Dispute Resolution Process

The next thing that I think was long awaited from the agency was the dispute resolution process. I do think that the dispute resolution processes have existed for quite a while, informal and formal. But what we were hearing is that industry felt that there wasn't an adequate avenue to raise scientific and technical issues and get a fair airing before the right agency experts. This guidance is intended to create a pathway to establish just that. It establishes procedures to resolve disputes about scientific and technical issues concerning GMP inspectional observations.

It is intended to do a few things in addition: l To enhance field/center communication and coordination - by getting the centers involved with the field on these scientific and technical issues that are raised l to facilitate the involvement of the most knowledgeable scientific and technical experts. In some cases that will mean getting the center product specialists and/or compliance staff involved l and also to contribute to the guidance development process. By creating a pathway to air and consider these scientific and technical issues, we will have a better idea where clarification is needed in the form of guidance documents to achieve greater consistency and understanding amongst our investigators and within industry.

It encourages informal resolution, of course, during the inspection itself, and that includes consultation with agency technical experts as appropriate. And then it defines two tiers of dispute resolution. It talks about domestic drug manufacturers starting with the district office. Now there is a slightly different process described in there for Team Biologics inspections, and there is a slightly different procedure described in there for foreign inspections. But I am going to, for simplicity's sake, focus on the drug inspections, pre-approval and post-approval.

Tier 1: If the district office disagrees with the issue that is raised and the dispute that is raised, automatically it will go to the center for review. The center will have the responsibility for assembling the right experts to evaluate the scientific and technical issue and then get back to the district, work it out and give you a written response. Then we hope to share the knowledge - not just within the FDA so we can learn about it, but outside in industry so you can learn about what FDA's position on this is. To do that, we will have to redact the confidential, commercial and trade secret information, if any, and share that on our internet site so that it is freely available.

Now if that doesn't work in Tier 1, Tier 2 will provide an opportunity to appeal unfavorable decisions to an agency level dispute resolution panel, which will consist of experts from each of the centers and additional experts as needed - if there are particular technical or scientific issues that need to be addressed, with appropriate representation. Again, a written response will be provided and a redacted version will be made public.

Now this is just a draft guidance, but in order to be able to evaluate and figure out how to tinker with, adjust and finalize this guidance, this approach, we are going to do a pilot. It is going to be a domestic pilot and it is going to be broad. I can't tell you which and how many districts at this time will be involved in it, but it is not just going to be one or two. That will start no later than January 1 of '04. It will be a 12 month pilot. We hope that this will give you a basis to send it in your comments. Rather than just looking at the guidance, the words or the guidance document itself, you will be able to comment on the guidance document as it is actually being used in the pilot. We think therefore the comments we get will be more productive.

Aseptic Processing

The next hot item we announced last week was the aseptic processing guidance. This is obviously a very important issue to this group, and I believe [FDA compliance officials]Rick Friedman and Brenda Uratani spoke about that earlier to you. So I am not going to go into any detail, but I do want to hit you with the highlights as I see it.

The draft guidance was issued by CDER, CBER and ORA. It updates a 1987 guideline which I think was long overdue for updating. One of the things that is important about this approach is that it takes a step toward harmonizing CDER and CBER's approach, which is important I believe for sites and facilities that may manufacture sterile products that are under both CBER and CDER regulation.

It enhances and updates communication and thereby facilities improved compliance in general. Modernization and automation are key themes. And I think the risk-based approach is another key theme. Identifying critical control points and addressing them are themes that run throughout the document.

The process involved extensive outreach to academia and industry. A concept paper was used to begin forming a dialogue. We got a number of comments on that. We had pharmaceutical science advisory committee meetings and extensive work with the Product Quality Research Institute, which was very helpful in resolving a number of the issues and doing data-mining to address scientifically some of the controversial and difficult issues.

It proactively promotes GMP compliance through enhanced understanding and consistency. Some of the new topics covered include personnel, isolator technology, blow/fill/seal processing, and early processing, and that includes some of the upstream activities that are particularly important I believe for biologics manufacturing. It takes a step toward harmonizing with EU standards, at least with regard to environmental monitoring expectations. It also includes an updated media fill section covering acceptance criteria and study design. And it acknowledges importantly the great advances of rapid micro test methods, and I believe encourages adoption of that technology.

That is only a draft guidance obviously, and this is an enormously challenging and complex topic. From this group in particular, we are depending on getting extensive and helpful comments. Because we need all of your input to make that final guidance as good as it can be.

Pre-approval Inspection Priorities

The next item that was announced last week concerns pre-approval inspection priorities. And I think this was perhaps a first step of many that we intend to take to provide greater flexibility to the field for determining when a pre-approval inspection is warranted - to give the field greater flexibility to exercise a risk-based approach, and to provide more resources for higher risk or higher priority pre-approval and post-approval inspections.

The step taken here involves modifications to the compliance program that the field uses and take away some categories that previously automatically trigger a pre-approval inspection no matter what FDA knew about that facility. Now there are many cases where we have adequate information from recent comprehensive systematic GMP inspections to know enough about the GMP state and the quality control and quality systems in place to dispense with a pre-approval inspection. Because of our limited resources, we need to take advantage of that knowledge and vest that discretion in the field. As I mentioned, this is the first step and you will be hearing more about this….

Comparability Protocols

Comparability protocols also were another announcement last week. In this case we issued a draft guidance for the use of comparability protocols for protein drugs and biological products. Now six months earlier we issued another draft guidance that concerned the non-protein, the small molecule drugs.

Both I think try to accomplish similar objectives - many of the themes of the GMP objective. They allow for certain manufacturing changes without submission of prior-approval supplements. And I think they facilitate continuous improvement and innovation. They target regulatory review according to the risk that a manufacturing change will adversely affect product quality consistent with the risk-based theme in the initiative….We are actively considering other options for managing manufacturing changes without the need for prior FDA review.

Process Analytical Technology

PAT is something that obviously is central to this initiative. It is something that I am sure has come up many times over the course of this seminar. A draft guidance was issued that I think will be a big step forward in describing the regulatory framework that encourages the adoption of this innovative manufacturing and quality assurance technology. It includes a set of scientific principles and tools for supporting innovation, and it includes a new regulatory strategy for accomplishing and accommodating innovation.

We hope this guidance document will provide essential data for enhanced process understanding, which I think is essential to the risk-based approach that we need to take….

Manufacturing Science

There was also a progress report issued concerning manufacturing science from that particular workgroup emphasizing the opportunities to enhance the risk basis of regulatory decision making with knowledge acquired in different phases of manufacturing and development.

Risk-based Approach

Risk-based approaches were mentioned in the progress report we issued last week. One of the things that we continue to work on is applying a risk-based approach to our own priority setting, applying risk-based approaches to our selection of facilities for GMP inspections. And we hope this will help predict where inspections are most likely to have the greatest public health impact.

The model that we are developing and that we will be releasing in the coming months for comment involves risk factors relating to three different categories. We probably could have put these in a variety of different categories. I went to one speech and someone recommended that instead of three we use ten. We could do that, but for now we are looking at at least three. They are: l facility factors - that would include things such as compliance history l product factors - things specific or in some cases intrinsic to the product being manufactured at the facility; and l process factors. In some cases the process understanding and the robustness of that process, the quality systems in place, will be almost reverse risk factors or risk-mitigation factors as was alluded to earlier.

The agency is also reviewing our internal agency compliance programs to implement risk-based approaches. So we are applying a risk-based strategy not just to where we go but to what we look at when we get there. This will be an iterative process. But many of these documents are many years old and we can [go] through here and encourage the investigators to focus on what we know matters the most - what potentially raises the greatest risk at a particular facility. That is one of things we hope to accomplish with that adjustment.

We also signed a…m aterial transfer agreement with some academics from Georgetown University and Washington University in St. Louis to support a study being conducted regarding factors that lead to superior performance in pharmaceutical manufacturing. This will follow up on a similar, very comprehensive study that was done in the semiconductor industry with great success to identify factors in organizational behaviour and development that were correlated with the highest performance. We think that this study will facilitate development of our own risk models and help us refine some of our risk factors in focusing on what matters most.

Pharmaceutical Inspectorate

The pharmaceutical inspectorate also is I think a very important step forward that we were able to announce. ORA and CDER signed a memorandum of understanding establishing a specialized cadre of inspectors that will be extensively trained and will devote a majority of their time to complex or high-risk, high-priority pharmaceutical inspections, and that will include some preapproval inspections as well.

As you may know now sometimes the inspector/investigator who comes to your plant will be a national expert who has a great deal of experience. But there has been a lot of turnover in the agency, a lot of new people, and you may get someone who did a food processing plant or a device plant yesterday and is coming to your plant today. It is difficult for FDA to consistently train a larger force that does a wide variety of things. By concentrating a number of inspectors that will be specializing, it will also give us opportunities to achieve greater consistency and to achieve more intensive training.

Procedures for cGMP Warning Letters

What we want to review is something that actually was announced in February. As of March 1, CDER and CVM began reviewing warning letter proposals that came in from the field that were previously issued directly from the field. We hope that this will facilitate field/center communication - in fact, I know it already has - and also coordination on the implementation and the application of these.

We are planning and will begin working on an internal assessment shortly. One of the things that we hope to do is to look at warning letters that have come in over the last few months and identify the areas where the center had a different approach than the field - where some districts look at things maybe differently - and to try to reconcile those different approaches with internal directives, more training, and guidance documents.

International Cooperation

International cooperation has been a guiding principle of the initiative. I think at the International Conference on Harmonization in Brussels we made a big step forward in that direction. In our September report we said that we think that greater harmonization of international scientific standards applicable to [continued on last page]

AD PAGE HERE

[continued from p. 22] product quality regulation will promote technological innovation for enhanced public health protection. So this is key to the initiative.

In Brussels the steering committee adopted the following agreed vision: a harmonized pharmaceutical quality system applicable across the life cycle of the product, emphasizing an integrated approach to risk management and the science. I think that that vision is going to be very important in the steps we take in moving forward in a harmonized approach to pharmaceutical quality.

As part of that effort, two expert working groups were established as was mentioned earlier: one will focus on risk management principles; the other on quality by design in the pharmaceutical development process and throughout the lifecycle of the process.

Quality Management Systems

We also have been looking at quality management systems and quality systems throughout the initiative, both internally within FDA and externally in terms of the regulatory expectations we have with industry. We now have three different types of quality systems working groups that are set up to explore different aspects of our work in this area:

l The first is a "quality systems framework" workgroup. This is going to focus mostly on enhancing on our own internal quality systems - integrating those throughout the agency. l The second is the "quality systems guidance development" workgroup. This will focus on developing guidance documents that better articulate and amplify quality system principles that are really inherent and implicit in GMPs - which are less explicit in our drug GMPs than they are for example in the device quality systems approach. l We also created a third group that focuses on quality systems from the harmonization perspective. The "cGMP harmonization analysis" workgroup will analyze various internal and external CGMP requirements and approaches both within FDA - looking at HACCP and QSIT, quality systems approaches in devices - as well as international approaches to quality systems, and identify differences, consider options for incorporating best aspects and prioritizing harmonization opportunities as we move forward.

The success of this initiative really depends to a large degree on active participation from a wide range of stakeholders. That is why I am here today. That is why we are devoting a substantial amount of effort to outreach the scientific conferences domestically and abroad. We need your help. We need help from all of the stakeholders from industry and academia. I look forward to your questions and your comments.

www.FDAAdvisoryCommittee.com

Videoconferencing, Live and Archived Webcasts And/Or Videotapes Available for the Following FDA Advisory Committee For Pharmaceutical Science Meetings

• October 21-22 - FDA's Advisory Committee for Pharmaceutical Science will meet to discuss FDA's PAT draft guidance, and risk-based CMC and nomenclature.
• September 17 - FDA's ACPS Manufacturing Subcommittee met to discuss "quality by design" and to discuss and define risk management principles.

For information, visit www.FDAAdvisoryCommittee.com. For assistance, call Julie Robenson, 301.664.7203 or email [email protected].