Considering Annex 1 Revisions: Expert Insights

Published on: 
Pharmaceutical Technology, Pharmaceutical Technology, July 2021 Issue, Volume 45, Issue 7

The draft revisions of Annex 1 are driven by a quality risk management approach and will provide more clarity and detail for manufacturers.

The European Union’s good manufacturing practice (GMP) guide for sterile products, Annex 1, has been under review for some time, with the revisions being considered more along the lines of a rewrite due to the extensive and comprehensive nature of the changes. In a webinar, hosted by the Parenteral Drug Association (PDA) Ireland Chapter, experts outlined key changes, considerations, interpretations, and ways of implementing the revised guidelines (1). In this article, some of the fundamental insights from that meeting are summarized.

Driven by quality risk management

Pre-use, post-sterilization integrity testing (PUPSIT) is not a new requirement, explained Andrew Hopkins, director operation quality, Quality Assurance Audit and Compliance, AbbVie, ex-Medicines and Healthcare products Regulatory Agency (MHRA) regulator, and previous member of the Annex 1 working group. As there was no clear push from regulators or industry on the matter of PUPSIT previously, there is the misconception that it is a new addition brought in with the latest revisions; however, it has been included in the Annex 1 guidance since 1997, he revealed in the webinar.

PUPSIT specifically pertains to the integrity testing of the sterilizing filter and assembly. The reasoning behind why it is necessary is because filters can become “blind” and can mask flaws during post-use integrity testing, which would have an impact on the quality of a product. Work on the mechanisms and risks of filter flaw masking is being performed by PDA and the Biophorum Operations Group (BPOG); the results are not yet completed or published at the time of writing (2), but there has been some evidence of masking, which means it is an issue to be wary of, warned Hopkins.

In the latest revisions of Annex 1, however, the wording for PUPSIT is much more driven by quality risk management (QRM), which means that satellite navigation risk assessments cannot be used when risk assessments are performed. It is important to be completely honest about them and base them on scientific justification, emphasized Hopkins.

The contamination control strategy (CCS) is also based on QRM principles, with an emphasis on good design for equipment, facility, and processes in the first instance, then on well thought out procedures, and finally on

the monitoring system, Hopkins stated. The last sentence in the revised document, which says, “exclusively monitoring or testing does not assure sterility,” was highlighted as being of particular import by Hopkins (3).

Another aspect of the draft Annex 1 revision, that caught the attention of Hopkins, is the fact that there is discussion around how a CCS is a number of aspects that are put together rather than individual items, which would not achieve the required item in isolation. “So, it’s how all of those individual items work together to give you the overall CCS,” Hopkins added.

Furthermore, the revised guidance emphasizes the need for the correct people to develop the CCS (i.e., those people with the necessary technical and process knowledge). “I think that [aspect] is really important,” added Hopkins. “[The CCS] is not something that is developed by either the quality assurance (QA) manager, sat in his office, or the production manager, or one person, it is about making sure that all of the right people are involved.”

“The other key thing to realize,” Hopkins continued, “is that [the CCS] is supposed to be a dynamic process, and [the revised draft] talks to periodic reviews.” Based on his experience as a regulator, Hopkins’ opinion is that a periodic review tends to indicate an annual occurrence, at least.

Currently, the Annex 1 that is in place is “silent on the CCS” revealed Hopkins. However, this does not necessarily indicate that it is a new requirement, and, in fact, there was some talk of contamination control in the Chapter 3 GMP requirements that were brought into operation in 2015 (4), he added.

QRM principles, which are also discussed in the International Council for Harmonisation Q9 guideline (5), are important to bear in mind when approaching a CCS, asserted Hopkins. These principles are focused on scientific knowledge and protecting the patient, but also, are steering the industry to the fact that the level of effort put into the evaluation of risk to quality should be commensurate with the risk. “So, with steriles, we’re going to be tending toward that higher end of the risk element, but there may be bits that aren’t quite as important as other bits, so make sure you put the right work in at the right stage,” Hopkins confirmed.

The CCS needs to be a living document that is reviewed periodically, taking into consideration various feedback on processes, information on monitoring, or technological improvements that need to be included, and so on, to ensure that the strategy continues to work, Hopkins stressed. Importantly, the document should be written in a way that it is understandable to an inspector, and widening that out, it should be simple enough for everyone to understand, he summarized.


Aseptic process simulations

Despite there not being much change in terms of the Annex 1 revision and the requirements of aseptic process simulations (APS) for manufacturers, there is certainly now a lot more detail included in the guidance, revealed Greg McGurk, executive director Industrial Operations and Product Supply, Regeneron Pharmaceuticals, ex-Health Products Regulatory Agency (HPRA) regulator, and previous member of Annex 1 working group.

An aspect of the thought processes behind APS in the revised guidance is whether or not they are a true aseptic validation strategy or if they are an evaluation of the process and ongoing process controls, which led to the topic being moved in the guidance to the section where process monitoring is discussed, McGurk explained in the webinar. “When you consider [APS], they do form part of the overall CCS, so to speak, in that they are used to monitor those controls and minimize the risks associated with aseptic processes,” he said.

General aspects of the requirements laid out in the draft revision document are that manufacturers must represent the worst-case operating parameters in APS, which is nothing new, but the revision incorporates an element allowing for the use of surrogate materials, McGurk specified. “Typically, we haven’t seen the allowance of surrogate materials over the years,” he noted. “APS is a science-based approach, however, and so whatever material you are using should not inhibit growth and should not impact the capabilities to recover growth or contamination. So, if a surrogate material is used, it has to be based in good science.”

Processes should be imitated as closely as possible, with a requirement for all aseptic operations, manipulations, and interventions that routinely—and, in some cases, non-routinely—would occur, to be evaluated, McGurk stressed. “One aspect that has been incorporated into the revision is a section on more complex unit operations,” he said. The intent of the section was in relation to an attempt to achieve a continuous evaluation of processes from one unit to another. The guidance in the draft states that this should be avoided where possible, but under certain circumstances it can be justifiable, McGurk added.

Lyophilization evaluation expectations are more detailed in the revised guidance, McGurk highlighted. The loading, unloading, and chamber dwell time information are required, and manufacturers are required to replicate the maximum time between sterilization and lyophilization, as well as the maximum hold times post‑sterilization and lyophilization, he explained. All those aspects need to be incorporated into the overarching strategy, in addition to the largest number of trays that will need to be lowered into the lyophilizer, the longest duration of loading, the length of time the chamber may be open, and so on. “I think it is good that the draft does provide an additional degree of clarity around those expectations,” McGurk stated.

Additionally, the draft Annex 1 provides some definition around inherent and corrective interventions, McGurk continued. It is important to incorporate these interventions into the APS so they can be evaluated and addressed appropriately, but when designing or selecting the types of interventions to include, it is critical that the choice is not justified based on the facility design or process design, McGurk emphasized. “[Justification via facility/process design] is just not based on good science, it is not based on minimizing risk necessarily,” he said.

The draft Annex 1 provides some additional guidance on how to develop the APS strategy overall and includes the expectations of elements that must be considered in the strategy. “So, obviously it is necessary to identify and assess the worst-case container size, line speed, container closer, or all of those, and you have to look at all the variables,” McGurk specified. “If a bracket or matrix approach is being taken, then you should justify the choice of variables that are being incorporated into the overarching APS strategy.”

Volume and agitation are other considerations in the APS strategy. It is therefore important to ensure whatever volume is being used can contact all the equipment in component surface, and if agitation and/or inversion may be required to ensure full exposure to surfaces is achieved, McGurk explained. A consideration has also been incorporated into the revised draft of Annex 1 for anaerobic simulation, he added.

The revised guidance thus provides more clarity on APS, and it is dependent upon the application of QRM to evaluate what is necessary to ensure a robust assessment of the process is being performed, McGurk confirmed. Furthermore, there is mention of manual filling, which is high risk and should be avoided where possible, McGurk added. Again, the revised draft provides guidance and clarity on the expectations for such processes that are high risk.

Moreover, it should be noted that the target for APS is to have zero growth, so any contaminated unit will result in a failed process simulation, McGurk stressed. “There were some caveats in the Annex 1 previously, which allowed for a degree of failure,” he said. “However, when dealing with APS, it is imperative to understand what the failures are and to investigate the root causes of the failures.”

In determining the root cause of a failure, if a corrective and preventive action that impacts the APS is identified, then this will call into question every batch since the last successful simulation, and it may be necessary to perform the three consecutive runs again until success is achieved, McGurk stated. “So, there are multiple elements to incorporate within such investigations, and QRM is very much one of the key fundamental elements incorporated into the draft,” he summarized.


1. PDA Ireland Chapter, “Annex 1: Interpretation and Implementation of the Latest Revision,” Webinar, 11 March 2021.
2. T. Morris, et al., “PUPSIT and the Annex 1 Revision,”, 29 Aug. 2019.
3. EC, “Annex 1: Manufacture of Sterile Products,” Draft Guideline under Consultation, [Accessed 23 June 2021].
4. EC, EU Guidelines for Good Manufacturing Practice for Medicinal Products for Human and Veterinary Use. Part 1. Chapter 3: Premises and Equipment (Brussels, 13 Aug. 2014).
5. ICH, Q9 Quality Risk Management, Step 4 version (2005).

About the Author

Felicity Thomas is the European editor for Pharmaceutical Technology Group.

Article Details

Pharmaceutical Technology Europe
Vol. 33, No. 7
July 2021
Pages: 39–41


When referring to this article, please cite it as F. Thomas, “Considering Annex 1 Revisions: Expert Insights” Pharmaceutical Technology Europe 33 (7) 2021.