The Annex 1 2022 Revision: A Success Story, Still to be Written

The Annex 1: Manufacture of Sterile Medicinal Products revision is a success story because, in our opinion, it is a good document and an improved version of its 2008 predecessor.

The revised Annex not only adds much-needed levels of detail and helpful content but also addresses more topics and is more relevant to product manufacturers and technology suppliers. The Annex presents the expanded expectations of European, PIC/S, and WHO regulators for the manufacture of sterile products, including those produced aseptically and those terminally sterilized. This may be the first time the same set of pharmaceutical requirements and guidance have been adopted by so many countries through EMA, PIC/S, and WHO participation. It can be viewed as guidance or regulatory requirement depending on how those jurisdictions, the countries they represent, and, to a certain extent, the users of Annex 1 view it (“twin” documents to the European Annex 1 have been issued by PIC/S in September 2022 and by WHO in October 2022 – with minor differences and with the same implementation timelines). Whether one agrees with all aspects of the document or not, it does present the thoughts of a collection of global inspectors and the thoughts of industry experts who we believe influenced those regulators. It should be seen as an important guide for the industry.

While the Annex 1 revision is a step in the right direction, the full success story still needs to be written because it is not without flaws. In our modern, complicated, sterile medicinal product business environment, perfection would certainly be helpful.

Positives and Concerns of the Revised Annex 1

There are some good things that we see in the updated Annex. The Annex:

• Has global applications and supports harmonization of requirements internationally. 
• Is an improvement on previous versions and other similar guidelines.
• Is risk-based and can allow for alternative approaches where those alternatives meet the intent of this Annex.
• Requires a documented review of the totality of controls in a company’s contamination control strategies (CCS).
• Reflects and addresses many of the comments and concerns of industry and builds on knowledge gained since the previous version was published.
• Provides helpful guidance for the manufacture of low-bioburden products and ATMPs.
• Provides more detail on regulatory expectations for modern manufacturing systems, including distinct sections and requirements for aseptic processing simulation (APS), environmental controls, personnel qualification, barrier technology, restricted access barrier systems (RABS), isolators, and blow-fill seal technology.
• Makes it clear that physical barriers are required to separate people from product, and it distinguishes requirements for RABS from isolators.
• Provides more detail on regulatory expectations for environmental monitoring, cleanroom qualifications, and aseptic process simulation, including the correction and clarification of acceptance criteria.

However, there are still questions and concerns that we still have, such as the following:

• In some instances, the language remains open to individual inspector interpretation, which can result in confusion, redundant effort, additional cost, lack of compliance with important points, and potential risk to product supply.

For example, it remains unclear if settle plates are required. For qualification purposes, Note 1 of Table 2 in the revised Annex 1 states, “All methods indicated for a specific grade in the table should be used for qualifying the area of that specific grade. If one of the methods tabulated is not used, or alternative methods are used, the approach taken should be appropriately justified.” To the contrary, however, Note 1 of Table 6 states, “It should be noted that the types of monitoring methods listed in the table above are examples and other methods can be used provided they meet the intent of providing information across the whole of the critical process where the product may be contaminated (e.g., aseptic line set-up, aseptic processing, filling, and lyophilizer loading).” So, does this mean settle plates can be replaced by other means (not easy for qualification, easier for monitoring)?

• It is still unclear whether inspectors will allow for alternate approaches based on an assessment of risk, as mentioned in paragraph 2.2 of the revised Annex 1, which states “… where alternative approaches are used, these should be supported by appropriate rationale, risk assessment and mitigation, and should meet the intent of this Annex.”

Take, for example, pre-use post-sterilization integrity testing (PUPSIT). The 2008 version of Annex 1 stated that “the integrity of the sterilized filter should be verified before use” but gave no more explanation of the purpose or reason for the test (according to paragraph 113 in the document). The revised version expands this recommendation, adding that PUPSIT should be performed “to check for damage and loss of integrity caused by the filter preparation prior to use.” The revised Annex goes on to recognize that PUPSIT may not always be possible after sterilization due to process constraints (e.g., the filtration of very small volumes of solution). It is unfortunate that the word possible and the wording small batch example were used. It would have been more useful to replace possible with feasible or practical and remove what will surely become a prescriptive example constraint, given the advancement of technology and related needs. However, the revisions did note that “an alternative approach may be taken providing that a thorough risk assessment has been performed and compliance is achieved by the implementation of appropriate controls to mitigate any risk....” It then goes on to list some guidance for what to consider in that risk assessment. However, it is disappointing that the most significant concern in the industry — the risk of contamination to the aseptic process by implementing PUPSIT — is omitted from that list.

Finally, the word should in the paragraph was encouraging because it insinuates that this is a recommendation rather than a strict requirement, or so we thought. This example of the use of the word should, when the intention may have been that the company must comply, is a concern throughout the Annex.

• The revised Annex 1 notes a requirement for sterilization of direct and indirect product contact parts (Section 5.5). It is undeniable that the surfaces of these parts should be rendered incapable of contaminating the product. However, a requirement for the sterilization of large items, such as stopper bowls in isolators, may prove to be problematic because of the concerns raised by some IWG members to vapor phase hydrogen peroxide (VPHP) as a sterilizing method. These concerns were noted in an April 20, 2018, MHRA on-line blog. The result may be that companies moving to a more acceptable, alternative sterilization method (e.g., by steam) will require more time than the prescribed period to implement.

• The Annex requires 100% (nonvisual) integrity testing of final product containers equal to or less than 100 mL and sealed by fusion. This presents a problem for the industry because there may not be technology available to test smaller-volume double-bagged products with multiple ports, such as those used for drugs and ATMPs, without damaging or weakening its integrity. This is especially of concern because container closer integrity testing (CCIT) of complex containers may stress the package, thus exposing it to failure afterward. Again, moving to an acceptable integrity-testing method will take longer than the prescribed implementation period. We question whether the manufacturers of complex or multiple-bag containers will be able to comply with this requirement or be able to develop the needed technology to do so.

• Paragraph 8.128 states that “…connection of sterile equipment (e.g., tubing/pipework) to the sterilized product pathway after the final sterilizing grade filter should be designed to be connected aseptically (e.g., by intrinsic sterile connection devices). Interpretations of this statement have raised concerns over whether tube welding in certain applications is considered to be an aseptic process that must be performed in a Grade A environment, as apparently interpreted by such agencies as Swissmedic (see sidebar) as to whether inspectors will allow for a risk-based assessment of tube welding in ATMP, vaccine, and other the manufacturing processes to show that the welding process is a closed system process that may not require the same level of control.

• As noted in the preceding examples, there appears to be too much emphasis on testing and monitoring, including APS and PUPSIT, as a means to qualify and prove the effectiveness of process control, rather than an emphasis on proper process control design. This is especially evident in sections related to personnel qualification and intervention control. The question remains whether the document can help companies move from reliance on process and product testing to process design as a primary means of controlling critical process risk.

• The implementation schedule is difficult to achieve and may lead to rushed efforts for compliance rather than a careful analysis of options. In some cases, it will involve the development of technology. Even where the technology exists, time will be needed to analyze and evaluate viable options, design implementation strategies, acquire and install equipment, train operators, and qualify these new systems. There is little guidance on how inspectors and the industry should react to delays in implementation, and what exists is not harmonized. The question remains as to how product manufacturers and inspectors will address delays regarding implementation of the revised Annex 1 requirements.