Media Fill for Powders and Lyophilized Products: APS in Isolators and RABS

Media fill for powders and lyophilized products is an aseptic process simulation (APS) that replaces the product with sterile nutrient medium, a surrogate, or both. Sterile powders need an acceptable surrogate in the same containers. Freeze-drying needs a simulation of the whole chain from filling to sealing, without freezing or boiling the medium. Isolators and RABS need their own transfers, glove work and campaign length built in. EU GMP Annex 1 sets zero growth as the target, and any contaminated unit fails the APS. FDA’s 2004 guidance uses graded criteria by run size instead.

What Is Media Fill for Powders and Lyophilized Products?

Media fill for powders and lyophilized products is an APS adapted to products that are dry at filling or that change during freeze-drying. Annex 1 paragraph 9.32 calls the APS a periodic verification of the effectiveness of aseptic controls. The paragraph adds that the aseptic process is not validated mainly through the APS. Process design and the pharmaceutical quality system (PQS) carry that role. So do controls, training and monitoring data.

For the full overview, see What is Media fill in Pharmaceutical. Here the focus is limited to the differences for powders, freeze-drying, isolators and RABS.

An isolator is an enclosure that separates the grade A zone from the room. A Restricted Access Barrier System (RABS) does the same job with a different design. Annex 1 paragraph 4.18 treats them as different technologies. You can read more in Defining a Pharmaceutical Isolator and Restricted Access Barrier System (RABS) in Pharmaceutical.

Related topic: media fill procedure and protocol

Why Does APS Design Differ From Liquid Filling?

In a liquid fill, the medium takes the place of the product and runs through the same pipes, pumps and needles. You cannot copy a powder or a freeze-dried product that way. Annex 1 9.32 says to pick the medium or surrogate by how well it imitates the physical product traits that pose a risk to sterility. Some stages may indirectly affect the survival of contaminants: powders, heating, cooling and lyophilization. For these, you develop alternative procedures that represent the operations as closely as possible.

Three rules in Annex 1 paragraph 9.33 matter most here:

• Sterile powders need an acceptable surrogate material in the same containers as the process (9.33(iv)).
• Avoid separate simulations of single unit operations (9.33(v)). Examples are drying, blending, milling or subdividing a sterile powder.
• The simulation for a lyophilized product covers filling, transport, loading, a representative chamber dwell, unloading and sealing (9.33(vi)).

How Do You Design the APS for Powder Filling?

Annex 1 names no surrogate material and no growth promotion method for powders. The site chooses the surrogate and records the reasons in its contamination control strategy (CCS), which is the documented set of controls against contamination.

Every choice has to meet three conditions. The surrogate goes into the same containers as the real process. It must not inhibit the growth of any contamination, which matters when a buffer is used (9.32). Its behaviour, such as flow, clumping or dust, should match the traits that put sterility at risk. A dry surrogate cannot show turbidity by itself, so the detection method must be scientifically justified (9.36(vii)).

For example, a team might run a powder fill, a separate liquid fill and a separate sealing check. Then it calls the three together a media fill. Each part passes, but nobody simulated the transfer between them. That is the gap 9.33(v) closes. If individual simulations are unavoidable, the record must show that together they cover the whole process.

For sterile active substances, 9.37 adds a rule on simulation materials. All of them, surrogate or growth medium, undergo microbial evaluation and must not compromise recovery.

How Is a Lyophilized Product Simulated?

The simulation follows the real chain from filling to sealing. It leaves out only what would harm the viability or recovery of contaminants. Annex 1 9.33(vii) gives boiling-over and actual freezing of the solution as examples to avoid. Paragraph 8.121 calls lyophilization a critical step and treats every activity that can affect sterility as an extension of aseptic processing.

A Example sequence looks like this:

1. Fill the medium and partially stopper the containers.
2. Transfer the partially closed containers to the lyophilizer under grade A conditions at all times, with minimal direct operator intervention (8.126(ii)).
3. Load by the specified and documented loading pattern (8.126(i)). Use worst-case parameters, such as the most trays and the longest time the chamber stays open to the environment (9.33(vii)).
4. Hold the chamber for a representative dwell. Break the vacuum with air instead of nitrogen (9.33(vii)).
5. Unload. Keep unsealed containers under grade A and use sterile utensils (8.126(iv) and (vi)).
6. Seal the containers, then incubate them.

The APS also replicates the longest allowed interval between lyophilizer sterilisation and use, and the longest time between filtration and lyophilization (8.122 and 9.33(vii)).

FDA’s 2004 guidance gives the chamber principle. Expose unsealed containers to partial evacuation that simulates the process, do not freeze the vials, and keep the medium aerobic.

Annex 1 sets no pressure, temperature or hold time for this step. Manufacturing Chemist (July 2024) describes one published approach: a partial vacuum of 800 to 900 mbar, containers at about 2 to 25 °C, stoppering after aeration with sterile air, and a hold of about 1 hour at atmospheric pressure so particles settle. A white paper in Pharmaceutical Online (2021) suggests three evacuation and aeration cycles at 100 to 200 mbar. The two differ, so treat both as examples of a justified site choice.

Sterilisation frequency shapes the design too. Under 8.123, you sterilise a lyophilizer before each load when it is loaded or unloaded manually with no barrier technology separation. For automated or closed-barrier loading, the site justifies and documents the frequency in the CCS. Paragraph 8.123 applied from 25 August 2024, a year after the rest of Annex 1.

Loading strategy is site practice. The Pharmaceutical Online white paper rates fully automatic loading as low risk. For that case, media-filled units plus enough stoppered empties to fill the last shelf may be enough. It rates manual loading as high risk and simulates the maximum commercial load. Annex 1 does not tier loading by automation. It requires worst-case conditions that you justify (9.33(vi)).

What Changes in an Isolator or RABS APS?

The APS must follow the technology’s own transfers, glove work, door openings and decontamination steps. Annex 1 paragraphs 4.19 to 4.22 set different rules for each technology.

Point	Isolator	RABS
Background	Open isolator generally grade C minimum, closed isolator grade D minimum, set by risk assessment (4.20)	Grade B minimum (4.20)
Airflow studies	At the interfaces of open isolators (4.20)	During interventions, including door openings where applicable (4.20)
Gloves	Leak test at least at the start and end of each batch or campaign; visual check at each use (4.21)	Sterilised before installation; sterilised or bio-decontaminated before each campaign; disinfected after exposure to the background; periodic integrity tests (4.21)
Bio-decontamination	Automated, validated, with a sporicidal agent (4.22)	Routine sporicidal disinfection by a validated method (4.22)

Paragraph 4.20 lists what an isolator risk assessment weighs. These are the bio-decontamination programme and the extent of automation. They also include glove manipulations that could compromise first air and loss of barrier or glove integrity. The last items are transfer mechanisms and set-up or maintenance that opens doors before the final bio-decontamination. Use that list to choose which transfers and interventions go into the APS.

Barrier lines often run in campaigns. Annex 1 9.36(xii) asks you to simulate the risks at both the start and the end of the campaign. You also show that the campaign length poses no risk. An end-of-production or end-of-campaign APS may add assurance or support an investigation. You justify it in the CCS, and it cannot replace the routine APS (9.36(xiii)).

Treat the APS like a production batch for glove testing. Test the gloves before and after the run, as 4.21 expects for isolators. That is site practice following 4.21, not a separate rule.

FDA’s 2004 guidance says an isolator process has low contamination risk because there is no direct human intervention. You can simulate an isolator process with fewer units as a proportion of the operation. Annex 1 9.40 still asks you to justify the unit count in the CCS.

How Does Risk Assessment Set Worst-Case Conditions?

Start with the process steps, then ask which variable stresses sterility most. Annex 1 9.36(i) requires worst-case conditions across the relevant variables, such as container size and line speed. The assessment must justify the choice. The same paragraph covers maximum holding times for sterile product and exposed equipment. It also covers extended shifts and idle interruptions such as shift changes. It adds a fill volume that wets every contact surface.

Paragraph 9.36(ii) allows a bracketing or matrix approach for container and closure sizes when process equivalence is scientifically justified. Interventions follow 9.34: inherent and corrective ones representative of routine production, at a similar manner and frequency, with inclusion based on assessed risk. Paragraph 9.35 adds that an APS must not justify practices that pose unnecessary contamination risk. For selection and timing, see Media Fill Interventions and Worst-Case Design.

Which Medium, Incubation and Limits Apply?

The medium must grow a designated group of pharmacopoeial reference organisms and suitably representative local isolates (9.36(vi)). FDA’s 2004 guidance names soybean casein digest medium as a general choice. For growth promotion, see SCDM, TSB and Growth Promotion Test in Media Fill.

After filling, agitate, swirl or invert the units so the medium reaches every inner surface (9.41). Incubate in clear containers (9.43). Annex 1 requires incubation conditions and duration that are scientifically justified and validated (9.44). FDA’s 2004 guidance sets 20 to 35 °C and at least 14 days. With two temperatures, it asks for at least 7 days at each, starting with the lower one.

After incubation, trained and qualified inspectors examine the units. A positive control with reference organisms and local isolates follows (9.45). Identify isolates to species level where practical (9.43).

An APS has limits. An APS cannot freeze the medium or pull a full vacuum on it. FDA also acknowledges technical limits on how precisely any simulation can characterise the controls. Where the real process cannot be copied, Annex 1 asks for alternative procedures that represent the operations as closely as possible (9.32). Document why each departure from the real process does not hide a contamination route.

What Are the Acceptance Criteria and Investigation Rules?

Annex 1 and FDA differ. Show both in the protocol, and state which one governs the site.

Source	Criterion
EU GMP Annex 1, 9.46	Zero growth is the target. Any contaminated unit is a failed APS.
FDA 2004, under 5,000 units	No contaminated units. One contaminated unit is cause for revalidation after investigation.
FDA 2004, 5,000 to 10,000 units	One contaminated unit: investigation, including a possible repeat. Two: revalidation after investigation.
FDA 2004, over 10,000 units	One contaminated unit: investigation. Two: revalidation after investigation.

The FDA criteria are nonbinding recommendations for FDA-regulated sites. FDA still says contamination at any run size may signal a sterility assurance problem. Do not invent a universal limit.

A failed APS under 9.46 triggers these actions:

• Investigate the most probable root cause and apply corrective measures.
• Run successful consecutive repeat APS, normally a minimum of 3.
• Review records since the last successful APS and assess the risk of sterile breaches in batches made since then.
• Quarantine product made on the line after the failure.
• Restrict operators until retrained and requalified, where the cause relates to operator activity.
• Resume production only after successful revalidation.

Every run needs full documentation (9.47). Reconcile units filled, incubated and not incubated, with a justification for any not incubated. Record each intervention with its start time, end time and the person involved. Put all microbial monitoring data in the APS batch record.

Units removed during the APS must match what routine production removes, never more (9.41). You simulate discarded material that touched product surfaces, such as a flush, with medium and incubate it (9.42). A run may be stopped early only if written procedures handle commercial lots in the same manner, and the stop needs a documented investigation (9.48).

Repeat the initial validation after a long period without operation (9.49). Repeat it after a change to process, equipment, procedures or environment. Repeat it also when a new container or closure is added. Link this to change control. Initial validation is at least three consecutive satisfactory simulations covering all shifts. Periodic repeats normally happen twice a year for each process, line and shift. Each operator takes part in at least one successful APS each year (9.38). For manual operations, each operator completes 3 consecutive successful APS and then one about every 6 months (9.39).

Which Guidance Applies?

Annex 1 of EU GMP Volume 4 has been in operation since 25 August 2023, apart from paragraph 8.123. PIC/S states that the revised Annex 1 of its GMP Guide (PE 009-17) entered into force the same day and is identical to the EU text with some minor differences. WHO’s TRS 1044 Annex 2 (2022) revises its 2011 sterile products guidance, prepared with the EU and PIC/S.

FDA’s guidance, dated September 2004, replaced the 1987 guideline. It quotes 21 CFR 211.113(b): you must follow written procedures to prevent microbiological contamination of drug products purporting to be sterile. These procedures must include validation of any sterilisation process.

Recent Technology Trends

• August 2024: Paragraph 8.123 took effect, so manually loaded lyophilizers without barrier separation are sterilised before each load (PIC/S and EU GMP Annex 1).
• July 2024: Manufacturing Chemist described post-Annex 1 lyophilizer APS practice built on partial-vacuum dwell, stoppering after aeration and tiered loading strategies.
• September 2025: A PharmaNow commentary reported sites moving to digital intervention logging and automated barrier systems. This is commentary, not an Annex 1 expectation.
• May 2026: PharmTech coverage described automated transport, loading and unloading in isolators as the main way to cut lyophilization risk. Annex 1 8.123 asks designers to minimise operator intervention as far as possible.
• September 2026: PIC/S issued PE 009-18, effective 24 September 2026. The change concerns Annex 19, not Annex 1.

Common Problems and Troubleshooting

Problem	Likely cause	Action
Contaminated unit	Intervention, transfer or loading step	Treat as a failed APS. Identify the organism and match the unit to time and activity. Follow 9.46.
Growth promotion fails	Medium or surrogate does not support growth	Investigate the origin of any contamination anyway and repeat the run promptly (FDA 2004).
Surrogate inhibits growth	Buffer or powder suppresses organisms	Replace the surrogate (9.32).
More units removed than routine	Clearance larger than the SOP states	Align the APS with the SOP. Never remove more than routine (9.41).
Glove fails leak test after the run	Glove damage during the APS	Investigate and assess the effect on the run and recent batches under the PQS.

Our Expert Tips

Frequently Asked Questions

Summary

An APS for powders and freeze-dried products must copy the real chain as closely as possible. Use a surrogate that does not inhibit growth, and avoid separate unit-operation runs. Isolators and RABS need their own interventions and campaign design. Annex 1 requires zero growth. FDA’s graded criteria apply to FDA-regulated sites only.

References

1. EU GMP Annex 1: Manufacture of Sterile Medicinal Products (2022), European Commission
2. Guidance for Industry: Sterile Drug Products Produced by Aseptic Processing, cGMP (September 2004), US FDA
3. PIC/S: Entry into force of revised Annex 1 of the PIC/S GMP Guide (PE 009-17), PIC/S
4. TRS 1044, Annex 2: WHO good manufacturing practices for sterile pharmaceutical products, WHO
5. Aseptic process simulations for lyophilisation: challenges and tips post-EU GMP Annex 1 revision, Manufacturing Chemist
6. A Better Approach To Aseptic Process Simulation For Lyophilized Products, Pharmaceutical Online
7. Advancing Aseptic Precision, PharmTech
8. Annex 1 Aseptic Process Simulation (Media Fill) design, PharmaNow