1What Annex 1 is and where it applies.
Annex 1 is the part of the EU guidelines on good manufacturing practice, EudraLex Volume 4, that covers the manufacture of sterile medicinal products. It gives technical guidance on the GMP principles in Directive (EU) 2017/1572 for human medicines, Directive 91/412/EEC for veterinary medicines and Delegated Regulation (EU) 2017/1569 for investigational medicinal products.¹ The current text, dated 22 August 2022 and published on 25 August 2022, came into operation on 25 August 2023, and point 8.123 on sterilising lyophilisers on 25 August 2024.¹² EudraLex lists it as “fully applicable since 25 August 2024”.³
It covers all sterile products, from active substances and excipients to finished dosage forms. Quality risk management applies to the whole annex, and its limits, frequencies and ranges “should be considered as a minimum requirement”.¹ The same text is Annex 1 of the PIC/S GMP Guide, adopted by the PIC/S Committee on 9 September 2022, and WHO revised its own sterile GMP in 2022 “in collaboration with the European Union” and PIC/S, as a harmonised text.⁴⁵
2A contamination control strategy across the facility.
Annex 1 puts design first: the facility, equipment and processes, then well-designed procedures, and last the monitoring that shows they work. “Monitoring or testing alone does not give assurance of sterility.”¹
A contamination control strategy (CCS) “should be implemented across the facility” to define all critical control points and assess the effectiveness of its design, procedural, technical and organisational controls and monitoring. It should be actively reviewed, drive continual improvement and form part of the periodic management review. Its 16 listed elements run from plant design, premises, personnel and utilities to vendor approval including single-use systems, outsourced activities, validation, cleaning, monitoring and CAPA.¹ Sterility test failures, environmental monitoring excursions and other non-conformities should be investigated before a batch is certified.¹
3Cleanroom grades and barrier technology.
Grade A is the critical zone for high-risk operations such as the filling zone, stopper bowls, open primary packaging and aseptic connections, normally under unidirectional airflow within a restricted access barrier system (RABS) or an isolator. Grade B is its background for aseptic preparation and filling, unless grade A is an isolator. Grades C and D are for less critical stages, for isolator backgrounds and for preparing terminally sterilised products.¹ RABS or isolators “should be considered in the CCS”, and any alternative should be justified.¹
Rooms are classified at rest and in operation, with sampling locations taken from ISO 14644-1. The maximum total particle counts, per cubic metre, are these.¹
| Grade | ≥0.5 µm at rest | ≥0.5 µm in operation | ≥5 µm at rest | ≥5 µm in operation |
|---|---|---|---|---|
| A | 3 520 | 3 520 | Not specified | Not specified |
| B | 3 520 | 352 000 | Not specified | 2 930 |
| C | 352 000 | 3 520 000 | 2 930 | 29 300 |
| D | 3 520 000 | Not predetermined | 29 300 | Not predetermined |
The manufacturer sets grade D limits in operation from a risk assessment, and classification with ≥5 µm particles in grades A and B may be considered where the CCS or historical trends indicate it.¹
4Sterilisation, filters and utilities.
“Where possible, finished product should be terminally sterilised”, since that gives greater assurance of sterility than sterile filtration or aseptic processing.¹ A sterilised filter assembly should be integrity tested before use (pre-use post-sterilisation integrity testing, or PUPSIT) and again after use. Where process constraints make PUPSIT impossible, as with very small volumes, an alternative needs a thorough risk assessment.¹
Glass ampoules, blow-fill-seal units and containers of 100 ml or less closed by fusion should be 100% integrity tested. Water for injections should be made by distillation or an equivalent process, which may include reverse osmosis. Lyophilisers loaded or unloaded by hand without barrier separation should be sterilised before each load.¹
5Monitoring, people and aseptic process simulation.
Grade A should be monitored continuously for ≥0.5 and ≥5 µm particles for the full duration of critical processing, at a flow rate of at least 28 litres a minute. The action limits for microbial contamination are these.¹
| Grade | Air sample (CFU/m³) | Settle plate, 90 mm (CFU/4 hours) | Contact plate, 55 mm (CFU/plate) | Glove print (CFU/glove) |
|---|---|---|---|---|
| A | No growth | No growth | No growth | No growth |
| B | 10 | 5 | 5 | 5 |
| C | 100 | 50 | 25 | Not set |
| D | 200 | 100 | 50 | Not set |
Any growth in grade A should lead to an investigation.¹ Aseptic gowning is assessed, visually and microbially, at least once a year, and unsupervised access to grade A and B areas is limited to people who have passed that assessment and taken part in a successful aseptic process simulation (APS).¹ Initial validation needs at least three consecutive satisfactory APS runs covering every shift. APS is then normally repeated twice a year for each aseptic process, filling line and shift, and each operator takes part in a successful APS at least once a year. Any contaminated unit fails the APS and calls for an investigation and normally at least three successful repeat runs.¹
6What changed in the 2022 revision.
The 2022 text revises the 2007 version of Annex 1, which had aligned the cleanroom classification table and added guidance on media simulations, bioburden monitoring and vial capping, coming into operation in 2009 and 2010.¹ The GMP/GDP Inspectors Working Group and the PIC/S Committee recommended the revision to reflect changes in regulatory and manufacturing environments, to make use of enhanced process understanding and the tools of ICH Q9 and Q10, to remove ambiguity and inconsistencies, and to take account of advances in technology.¹
PIC/S, EMA and WHO then formed a joint implementation working group in January 2024. Its recommendation PI 058-1, in force from 28 October 2026, answers key questions on the most important changes so that inspections are harmonised. Its stated purpose is “not to introduce any new GMP requirements”, and it records agreement that the technical interpretation of Annex 1 should be the same across PIC/S, the EU and WHO.⁴⁶
| Question | What PI 058-1 says⁴ |
|---|---|
| A site that cannot comply without rebuilding | Every requirement applies; assess the risk, set interim actions and a timeframe for compliance |
| How often to review the CCS | At least annually and at management review; no set format |
| ≥5 µm limits at classification | Not a default for grade A, or for grade B at rest; ≥5 µm monitoring in routine operation is required |
| PUPSIT | Required for filters sterilising the final product whenever feasible; “risk assessments should not be used inappropriately to justify the omission of PUPSIT” |
| Manual lyophiliser loading | Still allowed when the lyophiliser is sterilised before each cycle and safeguards are rigorous |
7How it runs in Seal.
In Seal, the contamination control strategy links contamination risks, their routes and barriers to the controls that address them and to the monitoring, qualification and investigation evidence behind each control. Every control claim shows its current evidence and owner, and a change cannot close until the affected CCS links and narrative are reviewed. The manufacturer defines the scientific strategy; Seal maintains the evidence network behind it.
For sterile injectables, the batch record is linked to environmental conditions, interventions, qualification, equipment state, materials, microbiology and release on one timeline, while line controllers and monitoring systems stay authoritative. In environmental monitoring, each sample keeps its location, collection window, media and incubation, so the collection time identifies which work overlapped. For aseptic process simulation, Seal holds the qualified process configuration and a coverage matrix maintained across runs, so a change or failure reopens only the coverage it affects.
References
- 1EudraLex Volume 4, Annex 1: Manufacture of Sterile Medicinal Products, C(2022) 5938 final, sections 1 to 9, European Commission, 22 August 2022. https:/
/ health. ec. europa. eu/ system/ files/ 2022- 08/ 20220825_ gmp- an1_ en_ 0. pdf (read 6 November 2026) - 2Revision, Manufacture of Sterile Medicinal Products, European Commission, 25 August 2022. https:/
/ health. ec. europa. eu/ latest- updates/ revision- manufacture- sterile- medicinal- products- 2022- 08- 25_ en (read 6 November 2026) - 3EudraLex Volume 4, Good Manufacturing Practice guidelines, European Commission. https:/
/ health. ec. europa. eu/ medicinal- products/ eudralex/ eudralex- volume- 4_ en (read 6 November 2026) - 4PI 058-1, Interpretation of the most important changes in the Annex 1 on the manufacture of sterile products (2022), PIC/S, 28 October 2026. https:/
/ picscheme. org/ docview/ 11495 (read 6 November 2026) - 5WHO good manufacturing practices for sterile pharmaceutical products, WHO Technical Report Series No. 1044, Annex 2, World Health Organization, 2022. https:/
/ cdn. who. int/ media/ docs/ default- source/ medicines/ norms- and- standards/ guidelines/ production/ trs1044- annex- 2- gmp- for- sterile- pharmaceutical- products. pdf (read 6 November 2026) - 6Adoption and entry into force of PIC/S Recommendation PI 058-1, PIC/S, 28 October 2026. https:/
/ picscheme. org/ en/ news/ 150 (read 6 November 2026)
AQuestions and answers
Does Annex 1 apply to non-sterile products?
Not as a requirement. Its scope is sterile products, but a manufacturer may apply principles such as the contamination control strategy, cleanroom classification, monitoring and gowning to non-sterile products where microbial, particulate and endotoxin control matters. It should then document which principles it applies and show that it complies with them.
Does the contamination control strategy have to be one document?
Not necessarily. There is no set requirement for its format. PIC/S, EMA and WHO expect a structured document in which each element is identified, described and evaluated, with a system for reviewing it regularly. Control systems that already work can be referenced rather than replaced, and the strategy should be reviewed at least once a year.
How often must aseptic process simulations be repeated?
Normally twice a year, about every six months, for each aseptic process, filling line and shift, and after any significant change to practices, facilities, services or equipment that affects sterility assurance. Each operator should take part in at least one successful simulation a year.
Is PUPSIT mandatory?
For filters that sterilise the final product, yes, whenever feasible. An alternative is acceptable only where process constraints not caused by design flaws make it impossible, justified in the contamination control strategy with a risk evaluation focused on product and patient safety. The filter is also integrity tested after use.
How does Annex 1 expect water for injections to be made?
By distillation or by a process equivalent to it, which may include reverse osmosis combined with electrodeionisation, ultrafiltration or nanofiltration. It should be stored and distributed so as to limit microbial growth, for example by constant circulation above 70°C.