The European Medicines Agency (EMA) kept its word, and the wait is over. While those involved with the manufacture of sterile products were assessing the text of Draft Revision 12 of Annex 1, the final approved EU GMP Annex 1: Manufacture of Sterile Medicinal Products was published 25 August 2022 (1). This put an end to speculation about what the final text would look like.
The third of the four 2022 PDA Annex 1 Workshops was held in Amsterdam 22–23 September. While the previous two events (in Dallas and Dublin) used Draft Revision 12 of Annex 1 as the reference point, this time participants could use the effective version.
[Editor’s Note: This article updates the author’s article of 4 Aug 2022 on the second PDA Annex 1 Workshop held prior to publication of the final Annex 1 guidance.]
Like the previous workshops, this event was characterized by a mix of presentations on specific Annex 1 topics combined with interactive workshops, whereby participants had the opportunity to ask the panel of experts about the presented subjects. Above all, it provided the opportunity to discuss the participant’s interpretation, challenges and possible solutions with peers from the industry. Based on the number of questions asked of the panel, this turned out to be a very successful workshop with a high level of interaction between participants and presenters and during round-table discussions.
The following six key takeaways could be extracted from the discussions and conversations.
Implementation Period and Readiness
With the publication of the final version of Annex 1, it became clear that companies need to comply with the requirements no later than 25 August 2023. An exception was made in the requirement for sterilization before each load of lyophilizers that are manually loaded and unloaded. Companies are expected to have “Section 8.123” in operation no later than 25 August 2024, an implementation period of two years following publication.
As regulatory representatives indicated during the previous PDA Annex 1 workshops, the final text was not expected to deviate much from the Draft Revision 12 version published in February 2020. When comparing the final published version with Draft Revision 12, it can indeed be concluded that the body of the text remained essentially unchanged with some points of further clarification added. Because of this, organizations had more than two years to start preparing and familiarizing themselves with the text and knew what was coming. That was one of the reasons EMA agreed to a deadline of one year for implementation, reemphasizing that the final published Annex 1 text is based on what regulators already have seen in the industry as a practice and, in essence, is not new.
Therefore, it was surprising to hear from different round-table sessions that there were still organizations who confirmed limited familiarity with the (Draft) Annex 1 within their company. Feedback received from multisite facilities indicated that they expect it will be difficult to implement all the requirements within the expected timeframe. For companies using contract manufacturing organizations (CMOs), there is an additional challenge as they depend on the readiness and willingness of the CMO to make changes. This could lead to serious financial costs and time constraints when it comes to regulatory approval of an updated Drug Master File.
Knowing that the clock is ticking, and implementation is expected with less than 11 months from publication of the final Annex 1 guidance, companies are urged to familiarize themselves with the final text, perform a gap assessment and create a realistic remediation plan with tangible actions to assure compliance beginning 25 August 2023.
Quality Risk Management
As we have seen, quality risk management (QRM) applies to the final Annex 1 document in its entirely, not only to specific paragraphs. The expectations are that QRM is to be used in a proactive and scientific approach, and data must be present to support the assumptions and decision-making process. QRM is not in place to “risk-out” good practices or expectations, nor to justify bad practices or work toward a predetermined outcome. When correctly used, QRM shows that an organization understands the characteristics of their products and proceeds to manufacture those products with facilities and utilities that support these activities. Potential risks are identified, and controls are in place to mitigate these risks. The point of a solid risk assessment is to prioritize and act on the identified risks. Besides the benefits for the company, it impresses regulators with how well the organization understands their process, how it is controlled and how potential issues are detected.
The maturity of QRM within companies varies. Some have risk management already embedded in their day-to-day operations while others are exploring how to implement the principles of QRM. It will be a challenge for the latter to add a structured, formalized risk assessment process for areas that were not included in previous Annex 1 drafts. Selection of the right risk tool is key as it will help to get the right answer. Often, companies default to the tools with which they are most familiar but may not necessarily be the optimal tool for their needs. It is like using a hammer to drive in a screw, it could work but it is not ideal. PDA and ANSI are currently collaborating and are soon to present an aseptic-processing risk assessment tool.
Good risk assessments are often led by a good, experienced risk facilitator who is familiar with the concepts and can help select the right risk assessment tool. This person guides the process with a multidisciplinary team and assures that the process is kept on track, as “scope creep” is a lurking threat. Where the responsibility lays for ownership and completion of risk assessments and who contributes and takes actions can be an equal challenge, potentially leading to delays or even risk assessment not being completed. For smaller organizations, this can be a challenge as it requires support from team members that could be needed elsewhere to keep processes going or resolve other issues. The use of an Understanding Responsibility Assignment Matrix (RACI Matrix) can provide a solution.
Contamination Control Strategy
The requirements for organizations to have their process and premises designed in such a manner that minimizes the risk for contamination are already laid down in Chapter 3 “Premises and Equipment,” Chapter 5 “Production” and Annex 15 “Qualification and Validation” and date back to 2015. The published Annex 1 reemphasizes these requirements with an expectation that each company has implemented a contamination control strategy (CCS) across the facility with specific focus on microbiological, pyrogen/endotoxin and particulate contamination. Annex 1 summarizes 16 elements to consider as a minimum that are to be part of the CCS. The document defines all critical control points present with an assessment of their effectiveness. It is a repository of elements and their control measures whereby it evaluates the end-to-end process in a holistic manner to prevent contamination. As the CCS is to be risk-based, CCS and QRM go hand in hand and are therefore closely related.
The points described in the QRM section above are also applicable for the development and maintenance of an effective CCS. Often, strategy documents describe what an organization is doing, but it does not rationalize why they are doing it as it is too often based on historical knowledge that was not written down.
The CCS must drive continuous improvement and, therefore, must be reevaluated periodically and adjusted where necessary. The change management and deviation process are drivers for reassessment of the strategy document. Trending of process, utility and environmental monitoring data also provide valuable information about the effectiveness of a company’s CCS and warrants periodic revisitation.
For CMO companies and their customers, it can be challenging to know what information to include in the CCS as some information could be proprietary for a product. Risks and proposed mitigation actions or changes are to be communicated to the client, whereby it is important to understand the risks associated with the process including the API, consumables and other materials. One solution could be to have one general CCS in place covering the CMO’s overall strategy augmented with a product- or process-specific CCS. This will enable customers to focus on their product-specific elements without the risk of sharing confidential information.
The PDA will soon publish a technical report for development of a CCS in pharmaceutical manufacturing that will provide guidance on how to establish such CCS in an effective manner.
Pre-Use Post-Sterilization Integrity Testing
Not surprisingly, pre-use post-sterilization integrity testing (PUPSIT) remained a highly debated topic during the PDA Annex 1 Workshop in Amsterdam. Although the U.S. FDA regulators are not enforcing PUPSIT, it will be the norm in and outside the European Union. There may be situations where PUPSIT can be omitted, such as for small-volume radioactives, but these must be rigorously scientifically justified with supporting. A rationale for not performing PUPSIT will be very difficult, and it could be a major compliance issue for companies who, to date, have not considered integration of PUPSIT in their final sterile-filtration process.
For processes that use single-use systems (SUS), it is expected to be designed in such a way that it allows for PUPSIT. Early engagement with the SUS supplier is paramount to ensure the system is designed correctly and they can meet the supply demand from a quality and quality perspective.
The PDA performed masking-test studies and demonstrated that some products are fouling and can clog damaged filters, leading to a passing post-use filter-integrity test. This strengthens the regulators’ position that it reduces the risk of using a nonintegral filter going undetected when using PUPSIT. As a former regulator righteously commented: it is not the question “how can we avoid PUPSIT,” as it should be the mindset of “how can we implement PUPSIT.”
Container Closure Integrity Testing
When comparing Draft Revision 12 with the effective revision of Annex 1, there is one adjustment regarding the requirement of 100% container closure integrity testing (CCIT) for units that are closed by fusion. While the Draft revision required that containers of all volumes be subjected to a 100% integrity test using a validated method, the final version was changed to include containers up to a volume of 100 mL. For larger volumes, the test frequency may be reduced and, if so, it must be scientifically justified and based on data. However, it is more important to focus on a robust process that prevents container closure integrity failures than to define a test strategy that confirms compliance of the product.
For units sealed by stopper and seals, integrity testing must be performed using a validated method, as visual inspection alone is not deemed sufficient. Apart from the initial validation and as part of product stability testing, not all companies perform routine CCIT on filled units. There are different technologies available that could be introduced inline or online, such as oxygen headspace measurement, vacuum leak tests, high voltage tests and residual seal force measurement, to name a few. There is no “silver bullet” or “one-size-fits-all” solution, and organizations are advised to seek support from expert companies. Like any other sampling program, the frequency must be based on risk and statistically justified. A lifecycle approach for vial and stopper filling seems to be the accepted practice.
CCIT verification must be part of transportation as transport conditions, such as decompression, can lead to the movement of plungers in prefilled syringes. Temperature fluctuations can affect integrity issues as experienced during the challenges with -80 °C transport-and-storage conditions of the mRNA COVID-19 vaccines.
Premises and Barrier Systems
Removing personnel from the open aseptic process is the most effective way to reduce the risk of introducing contamination caused by humans. This can be achieved by introducing a physical barrier between the critical Grade A zone in the form of a restricted access barrier system (RABS) or using isolator technology. Hence, the published Annex 1 puts a firm recommendation in place to consider the use of barrier technology.
Organizations must understand that there are differences in barrier technology including their controls. For example, using a RABS system, where doors are frequently opened to perform inherent or corrective interventions, are not deemed RABS but are deemed a classic Grade A/B design since it does not segregate operator interference from the critical zone during interventions. This was already spelled out in “USP <1116> Microbiological Control and Monitoring of Aseptic Processing Environments” (2).
Regarding using isolator technology, while the Annex 1 revision from 2009 required a minimum classification of Grade D as background for isolators, there are now further expectations. A Grade-D background is allowed when using a closed isolator system, that is, there is no infeed from or outfeeds to a lower classified area. The product, including all primary packaging components, are present in the isolator, and the entire aseptic process is completed, including sealing of the final product container within the isolator before the isolator is opened again.
Open isolators are those configurations where, for example, containers are transferred through a depyrogenation tunnel oven into the critical Grade A zone, units are filled and then moved out of the isolator via a mousehole. These require, at minimum, a Grade-C background and justification in the CCS.
When a product requires lyophilization, the design of the loading and unloading process plays an important role as it will dictate the sterilization frequency of the lyophilizer chamber. As per “Section 8.123,” manually loaded and unloaded lyophilization processes without barrier technology require sterilization of the lyophilizer chamber before each load. Companies have two years from the date of publication to be in compliance with this requirement. This could have implications for an organization’s manufacturing capacity and, therewith, a potential effect on market supply.
Retrofitting existing lyophilization set-ups can be a challenge, with the financial burden being only one aspect. In collaboration with engineering specialists, a workable solution must be found that allows for automatic loading and unloading of the current configuration or implementation of barrier technology. The available footprint can be a restricting factor, not to mention the delivery time for a reengineered process and the qualification and validation of the new process. These factors could easily exceed the allowed grace period of two years.
With the final text published on 25 August 2022, time is ticking for organizations to have the guidance in operation no later than 25 August 2023. With less than 10 months to go, companies must ensure they know the implications of the final version. A structured gap assessment and realistic remediation plan to address any noncompliances in a timely manner is important to satisfy regulatory inspections and “future proof” the company’s compliance with Annex 1. This requires in-depth knowledge of one’s manufacturing process and identification of associated risks augmented with remediation actions when it comes to correct adoption and successful implementation of QRM.
The final Annex 1 version calls organizations to adopt a new mindset that goes top-down, is holistic and may ask for cultural changes. Considerations are expected to be given to novel technologies, such as deployment of barrier technology, online CCIT and rapid/alternative monitoring methods. All geared toward the purpose of improving product quality and, ultimately, patient safety.
- European Commission (2022). Annex 1: Manufacture of Sterile Medicinal Products, The Rules Governing Medicinal Products in the European Union, Volume 4 EU Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use. https://health.ec.europa.eu/system/files/2022-08/20220825_gmp-an1_en_0.pdf.
- United States Pharmacopeia (2022). General Chapter ⟨1116⟩ Microbiological Control and Monitoring of Aseptic Processing Environments. In USP-NF. Rockville, MD: United States Pharmacopeia. https://doi.org/10.31003/USPNF_M99835_01_01.
About the Author
Patrick Nieuwenhuizen, Director Senior Consultant at PharmaLex, has over 25 years of experience in the pharmaceutical industry with a background in microbiology and sterile manufacturing. He has worked for several global pharmaceutical and biotechnology companies, including Genzyme (a Sanofi company) and MedImmune, across a variety of platforms from biologics to sterile fill finish and solid oral dose. Nieuwenhuizen has provided quality, sterility assurance and microbiology oversight for various site and laboratory expansion projects from construction design through to method transfer and operational readiness. Nieuwenhuizen has also been involved in several corporate initiatives, such as a sterility assurance council and the roll-out of corporate standard programs, designed to improve and maintain organizational quality standards. As a lead auditor, he has been involved with audits facing several competent authority inspections, including but not limited to Ireland’s Health Products Regulatory Authority, the U.S. FDA, ANVISA, the Chinese FDA and Health Canada. Acting as a risk facilitator for quality risk management programs, Nieuwenhuizen has gained significant experience with problem-solving and the management of complex investigations.
For more information on Annex 1, the author encourages readers to contact him: [email protected]