Features
September / October 2018

New & Improved: Meet the Annex 1 Revisions
Jean-François Duliere
Marick Paris-Cadet
Alexandra Yath
New & Improved: Meet the Annex 1 Revisions - Bottles

The European Commission (EC) published its long-awaited revision draft of Annex 1.


Disclaimer: This is a brief and informal synopsis of responses from UK MHRA, US FDA, and Russian State Institute of Drugs and Good Practices (SID & GP) regulators during a panel dialogue at the ISPE Europe Annual Conference in March 2018. It has not been vetted by any agency and does not represent official guidance or policy of the MHRA, FDA, or SID & GP. 

On 20 December 2017, the European Commission (EC) published its long-awaited revision draft of Annex 1: “Manufacture of Sterile Medicinal Products.”1 The Annex, part of the European Union good manufacturing practice (GMP) guidelines, has undergone several targeted updates since it was originally published in 1989; the last was in 2008. This is the first complete revision.2 ,3

In drafting the revision, the EC worked closely with the World Health Organization and PIC/S to maintain existing global standards; each organization will review the revised annex in a parallel public consultation.

Revisions and additions were significant, increasing the document from 16 to 50 pages. One of the most notable changes is the inclusion of quality risk management (QRM) principles, which are used to reduce the risk of contamination and maintain the quality of a medicinal product throughout the product life cycle. QRM, described in ICH Q9 and used in GMP since 2013, evaluates the manufacturing process and equipment to identify patient risk as well as document mitigation and management processes commensurate with that level of risk. While QRM is not specific to sterile products, its application is at the heart of the new Annex 1.

Additional changes include:

  • New sections: scope, utilities, environmental monitoring, process monitoring, and glossary
  • Reorganized and restructured content for a more logical flow
  • Introduction of QRM principles
  • Existing sections enhanced and expanded for better clarity

After the revised draft was released, targeted stakeholders from industry and national competent authorities were invited to consult and comment on the revision from 20 December 2017 to 20 March 2018.3

ISPE Comments

ISPE members submitted more than 700 comments on all of the document’s 11 chapters. After consolidation by SMEs and a final review by the Regulatory Steering Committee, 290 ISPE comments were submitted to the EMA.5 These were also presented at during the 2018 Europe Annual Conference in Rome.

ISPE’s Annex 1 comments and responses were also presented during the 2018 Europe Annual Conference in Rome, followed by a panel discussion with three regulators—Andrew Hopkins from MHRA, Rick Friedman from FDA, and Vladislav Shestakov from SID & GP. This Q&A session provided a lot of information and gave participants a chance to raise many additional questions and comments. While the discussion covered the entire annex, QRM was a frequent topic of discussion, as expected.

Scope

The following statement was the focus of much attention:

However some of the principles and guidance, such as contamination control strategy, room qualification, classification, monitoring and gowning, may be used to support the manufacture of other products that are not intended to be sterile (such as certain liquids, creams, ointments and low bioburden biological intermediates) but where the control of microbial, particulate and pyrogen contamination, to reduce it as far as possible, is considered important.

This raises industry concerns that all points of Annex 1 would become compulsory for nonsterile products.

Regulators stated that this reference to nonsterile products was included because there were no other references in other chapters or annexes about room classification (Grades A to D), gowning requirements, or contamination control strategy, even though those principles are not specific to sterile products. Though the inspectors understood the possible confusion and will try to clarify this point, it deals with products that require thorough controls and have low contamination limits.

Regulators also re-stressed the use of QRM principles to determine which points of Annex 1 should be applicable to nonsterile products. 

Personnel

This chapter deals with personnel, and although it is a small section, it generated questions as well.

Annex 1 states that microbial monitoring of personnel in Grade A and B rooms should be done “upon each exit from the cleanroom.” While this was not stated in the previous version of the annex, it was clearly desirable even before the revision, because manufacturers should know the contamination risk for and from operators in their environments. Monitoring provides valuable data, so it is recommended as much as possible.

The draft also introduced a new concept: “personal disqualification.” While personal qualification is fully integrated into GMP and is mandatory before entering a cleanroom, the new term “disqualification” was unclear to many participants. When is someone disqualified, and when is requalification necessary? Is a person who was sick for 2 to 3 days disqualified in the same way as someone who had a serious illness and was absent for 3 months?

Regulators explained that this statement was intended to cover trained operators who demonstrate unacceptably high contamination levels that harm product integrity. In cases like this, disqualification and requalification are mandatory. Operators with a cold, flu, or other mild illnesses are temporarily barred from entering airlocks and cleanrooms, but do not need requalification once they recover. Those returning from long illnesses require requalification, if only because after a protracted absence they will not be familiar with the process or able to work in alignment with GMP. 

Premises

Another change with potentially profound effects is the stated preference for separate cleanroom entry and exit airlocks. Many participants felt that the new text was unclear, and asked if the revision required separate airlocks for both new and existing facilities. Regulators explained, however, that this requirement currently applies only to new facilities.

A single airlock combined with a contamination control strategy (cross-contamination control) is acceptable for existing facilities. Nevertheless, regulators indicated that upgrading facilities by implementing barrier technologies such as isolators, RABS, or separate entrances and exits can reduce possible cross-contamination and provide better control.

While upgrading facilities to adapt to new guidelines can be troublesome, it’s an important component of patient safety and product quality. As such, it is industry’s responsibility to adapt to new regulations. Old facilities cannot ensure the quality required from new guidelines as they were not built to these standards.

With regard to the cost of upgrading, regulators acknowledge that implementing barrier systems can be a financial challenge, but they pointed out that supporting facilities built before 1990 is as costly as upgrading them. Regulators indicated they would study contamination risks for both old and new facilities.

While expensive, installing isolators can often reduce the grade of clean-room needed; the background required is at least a Grade D. Nevertheless, background grade must be determined by each individual case and risk assessment. Loading and unloading a lyophilizer where the stopper is not fully inserted, for example, makes contamination more likely to occur, and only grade A would be acceptable. A negative pressure isolator would also need a Grade A background.  

Grade A or B is required for all operations prior to containment without a subsequent sterilization process, since vaporized hydrogen peroxide (VHP) is no longer considered a sterilizing agent, although regulators allow it to be used as a disinfectant. Liquid peroxide solution removes bacterial contaminants more reliably with enough contact time on every surface (i.e., dipping). Exceptions can be proved if they use QRM and demonstrate a 6-log reduction.

Cleanroom classification also underwent a significant change in this draft: While 0.5 micrometer (µm) particle size is still used for room qualification, 5 µm particles, while not relevant for qualification, are now used for monitoring, as they tend to be the first indicators of a problem.

Additional discussions covered HVAC filters, environment background, and air flow visualization.4

Utilities

The Annex 1 revision includes the EMA Q&A document on production of water for injection (WFI),6 although WFI requirements are still not well developed. While the draft now states that WFI must be produced from purified water (previous versions required only drinking water), GMPs must be harmonized with the European Pharmacopoeia monograph 169. The EU Pharmacopoeia also recommends the use of reverse osmosis to produce WFI, but regulators prefer distillation. For the time being, the Q&A document is still the reference for producing WFI using reverse osmosis.

Production

Chapter 8, “Production and Specific Technologies,” is the most substantial chapter in the draft:  Not surprisingly, it received the most comments and questions.

To decide whether the correlation between stopper height and microbial ingress is acceptable in experimental data, for example, scientific experiments must use QRM to prove their ability to limit ingress and containment. Another example is using risk analysis in filter integrity tests.

During this part of the panel discussion, regulators provided some clarification on the pre-use post sterilization integrity test, or PUPSIT. EMA regulators appear to strongly favor its use. Hence, to justify rejecting PUPSIT in Europe, manufacturers will have to prove their case using QRM principles. Some manufacturers say that PUPSIT is difficult to implement, but regulators strongly disagree. A. Hopkins stated that he managed to implement PUPSIT 30 years ago when he was working in industry. It should be much less challenging now, he said, since technologies have progressed and better systems have been developed.

Other comments in this section centered on definition and wording clarifications, sterilization operations, autoclaves, and freeze dryers.4

Language

Another issue discussed during the session was the prescriptive language (“shall,” “should,” “must”) used throughout the document. Participants found it difficult to determine whether such statements were requirements or advice. In a regulatory document like Annex 1, clarity in terminology is important, especially for non-native-English speakers. In addition, the document should not be overly prescriptive, since QRM is not mandatory for the entire manufacturing process, although it is highly recommended for some stages. Other language-based comments included requests to improve the glossary, clarify definitions, and distinguish laminar from unidirectional flow.4

Q&A with Jean-François Dulière

Pharmaceutical Engineering magazine caught up with Annex 1 Comment Team Lead Jean-François Dulière to discuss the Annex 1 revision ISPE member comments. 

What Is Annex 1 and Why Is It Important?

Annex 1 is a GMP document published by the European Commission and dedicated to the manufacturing of sterile products. All pharmaceutical companies that produce sterile products must comply with these regulations. The document is a joint initiative between the EMA and PIC/S.

What Are the Major Changes?

The last revision to Annex 1 in 2008 contained 123 items. This revision has produced a completely new document, reorganized in alignment with GMP in Europe. With 269 major points, it addresses significantly more details. This revision is not as stringent as the pharmaceutical industry had expected and is based on quality risk management. This is critically important for the pharmaceutical industry.

Who Will Be Most Affected by This Revision If or When It Is Put into Force?

All companies that manufacture sterile products. Some parts also apply to products that require clean air or a clean manufacturing environment.

Does the Revised Annex Differ from Other Agency Guidances?

In some instances, yes. The goal was to create a consensus document with all PIC/S parties, and input from greater than 70 different countries was received. Not all differences were resolved, however.

How Serious Are These Differences?

The most significant was on the requirement for PUPSIT—pre-use, post-sterilization integrity testing—of filters. European regulators believe that this should be a requirement, while those in the United States do not. Many companies think that PUPSIT increases risk to the product because it may damage the filters.

What Were Some Common Concerns Expressed by ISPE Members?

Common concerns included questions surrounding PUPSIT and clarification of facilities design, as well as the use and operation of isolators and restricted access barrier systems.

What Were the Most Significant?

Questions and concerns surrounding PUPSIT were the most significant. Extensive use of QRM was also a concern for industry, particularly how their QRM practices will be received by regulators.

What Did You Find Most Surprising or Interesting about the Comments?

Industry stakeholders asked for clarification of quality risk management guidelines. This was surprising because these have been in use for many years. However, clarification is important because without clarification, regulators may assume greater flexibility to accept or reject industry QRM practices.

Did You See Any Regional Differences in the Members’ Concerns?

There was not much regional difference in the comments regarding the key points. This is likely due to the globalized nature of the pharmaceutical industry.

Is There Evidence That Science and Risk-Management Principles Were Applied during Development of the Annex?

Yes. One example is the approach to particle-size monitoring. In this version of Annex 1, regulators require sterile product manufacturers to monitor the size of particles present in cleanrooms. Standard classification practices and qualification will be done with 0.5  µm particles. If larger particles are detected, this is considered an early indicator of potential problems. The presence of these particles alone would not call for a batch to be rejected but should initiate a risk assessment process to prevent other problems in the system.

How Did the Comment Lead Team Reconcile so Many Comments in the Allotted Time?

The commenting period lasted from 20 December 2017 to 8 February 2018. The comments were then reviewed by a team of 18 people divided into seven groups based on subject-matter expertise. From 12 February to 9 March each group reviewed the comments for a specific chapter. I met several times with each subject matter expert team. At the conclusion of this period, all of the accepted comments were combined and reviewed by the entire Comment Lead Team. We then sent the document to the Regulatory Steering Committee, which did the final review.

Has There Been Any Early Response from Regulators to Industry’s Reaction?

Early responses came during the panel discussion at the EU Annual Conference in Rome. There was some surprise regarding industry’s request for clarification on quality risk management; as mentioned earlier, these guidelines have been in place for many years. There has also been some discussion around PUPSIT, with European regulators explaining why they want to maintain it, and US regulators saying that the decision should be based on quality risk management. There was also some discussion surrounding best practices for sterilization and decontamination. Despite industry support for vaporized hydrogen peroxide, regulators may require other methods. 

Has EMA Indicated Whether the Next Version of Annex 1 Will Be Final, or Is It Likely That Another Draft Will Be Issued for Comment?

Normally, a new draft would be issued by September or October, but the resources to do that may not be available. The EMA’s goal is to have a final document issued by the end of 2018.

Was the Us FDA Requested to Comment on Annex 1? Is It Possible That the FDA Could Adopt the Final Version of Annex 1?

The FDA has issued comments on Annex 1. It is not likely to become law in the United States but may become a guidance for industry.

Once Annex 1 Is Finalized, Will There Be a Transition Period for Its Adoption by Industry in Europe?

Normally, the adoption period would be six months. In this case, however, it is likely to be longer because of the significant number of details in the document. We could potentially see industry allowed one to two years for implementation for topics that require a long period for implementation in existing facilities

In the previous revision of Annex 1, which was finalized in March 2008, certain points were not required to be implemented until March of 2010 because they required facilities modifications and industry needed time to do this.

Any Further Thoughts about Annex 1?

In addition to the major points already discussed—PUPSIT, particle size, quality risk management, facility design, sterilization, and decontamination procedures—there were also some comments around how to qualify facility personnel. Overall, this latest version of Annex 1 is a significant improvement, as it provides significantly more detail, is better organized, and is based on quality risk management. 

Single-Use Integrity Testing

The panel dialogue ended with a question on single-use integrity testing. Because single-use systems appear more reliable, their integrity can be overestimated. Regulators cautioned that the ability to detect integrity failure does not guarantee a leak-free, contamination-free system. Single-use systems, therefore, still need integrity testing and do not necessarily remove operators from the process.

The goal was to create a consensus document with all PIC/S parties, and input from greater than 70 different countries was received

Has EMA indicated whether the next version of Annex 1 will be final, or is it likely that another draft will be issued for comment?
Normally, a new draft would be issued by September or October, but the resources to do that may not be available. The EMA’s goal is to have a final document issued by the end of 2018.

Was the US FDA requested to comment on Annex 1? Is it possible that the FDA could adopt the final version of Annex 1?
The FDA has issued comments on Annex 1. It is not likely to become law in the United States but may become a guidance for industry.

Once Annex 1 is finalized, will there be a transition period for its adoption by industry in Europe?
Normally, the adoption period would be six months. In this case, however, it is likely to be longer because of the significant number of details in the document. We could potentially see industry allowed one to two years for implementation for topics that require a long period for implementation in existing facilities

In the previous revision of Annex 1, which was finalized in March 2008, certain points were not required to be implemented until March of 2010 because they required facilities modifications and industry needed time to do this.

Any further thoughts about Annex 1?

In addition to the major points already discussed—PUPSIT, particle size, quality risk management, facility design, sterilization, and decontamination procedures—there were also some comments around how to qualify facility personnel. Overall, this latest version of Annex 1 is a significant improvement, as it provides significantly more detail, is better organized, and is based on quality risk management.

—Emily Burke, PhD

Closing

As the session concluded, Hopkins explained that with the comment period over, EMA will now review and assess all comments received to determine which should be included in the final version.4 The final release of Annex 1 is expected in December 2018.