1What a media fill is.
A media fill, which EU GMP Annex 1 calls an aseptic process simulation (APS), runs the aseptic process with a sterile nutrient medium, or a surrogate, in place of the product. The filled containers are incubated, and any growth shows that contamination got in. Annex 1 defines an APS as “a simulation of the entire aseptic manufacturing process in order to verify the capability of the process to assure product sterility”, from equipment assembly and formulation to filling, lyophilisation and sealing.¹
In the United States, 21 CFR 211.113(b) requires written procedures to prevent microbiological contamination of sterile drugs, including “validation of all aseptic and sterilization processes”.² FDA’s 2004 aseptic processing guidance says an aseptic operation “should be validated using a microbiological growth medium in place of the product”, exposing the medium to the same product contact surfaces, container closures, environments and manipulations the product meets.³
Annex 1 is clear about the limits of the test. The APS “should not be considered as the primary means to validate the aseptic process”. That rests on process design, the quality system, process controls, training and monitoring data. And an APS “should not be used to justify practices that pose unnecessary contamination risks.”¹
2How often it must be run.
Both regulators ask for at least three consecutive successful runs to qualify a line, then a repeat about every six months. Annex 1 counts the repeats per shift as well as per line.¹³
| Point | EU GMP Annex 1 (2022) | FDA guidance (2004) |
|---|---|---|
| Initial validation | At least three consecutive satisfactory simulations covering all working shifts the process may run in (9.38) | At least three consecutive separate successful runs during initial line qualification |
| Routine revalidation | “Normally”, twice a year, approximately every six months, for each aseptic process, each filling line and each shift (9.38) | Semi-annually for each processing line, with activities and interventions representative of each shift and shift changeover |
| People | Each operator takes part in at least one successful APS a year (9.38) | Everyone authorised to enter the aseptic processing room during manufacturing, technicians and maintenance staff included, takes part at least once a year |
| Manual operations | Each operator validated with at least three consecutive successful APS, then one about every six months (9.39) | Run duration generally no shorter than the actual manufacturing process |
| Changes | After significant modifications affecting sterility assurance, such as HVAC, equipment, process, shifts or staffing changes or a major shutdown (9.38) | Changes or events that could affect the process are assessed through additional media fills |
Annex 1 also asks for the initial validation to be repeated when a process has not run for an extended period, or after a change to the process, equipment, procedures or environment, or a new container closure (9.49).¹ The PIC/S, EMA and WHO interpretation in force from 28 October 2026 says that if an APS can’t be run twice a year for each shift and filling line, that is an extended period “by definition”, and that a line left unused without its routine APS programme maintained would be expected to repeat an initial APS of three runs.⁴
People are qualified through the APS too. Unsupervised access to grade A and B areas is limited to staff who have passed the gowning assessment and “have participated in a successful APS” (7.4).¹
3Isolators and closed systems.
FDA’s CGMP questions and answers allow a risk-based frequency. For an isolator line run on two shifts, “it may be justified to perform fewer than four media fill runs per year”, provided each aseptic processing line is still evaluated at least semi-annually. The lower number is subject to re-evaluation after product nonsterility, a media fill failure or adverse environmental trends, and the answer does not apply to RABS.⁵
The European reading is firmer. Asked whether a validated isolator system could run one APS a year by following ISO 13408-1, PI 058-1 answers: “No, ISO 13408-1 does not override Annex 1 requirements.”⁴
4How a run is designed.
A media fill should imitate the routine process as closely as possible, with its worst cases.¹³
- Interventions. Inherent and corrective interventions are performed in the manner and frequency of the routine process, selected by their risk to sterility.
- Duration. Long enough to challenge the process, the operators, shift changes and the environment. PI 058-1 asks for the whole filling time to be covered at initial validation.⁴
- Size. Enough units to simulate every representative activity, with the number justified in the contamination control strategy. Typically at least 5,000 to 10,000; for batches under 5,000 units, at least the size of the production batch (9.40).
- Units. Every integral unit is incubated, cosmetic defects included. No more units may be removed for an intervention than a production run would clear.
- Media and incubation. FDA names soybean casein digest medium and recommends at least 14 days at 20 to 35 °C, with at least 7 days at each temperature if two are used, lower first. Annex 1 asks for conditions that are “scientifically justified and validated”.
- Records. A batch record for each run, with every intervention and its start and end time, the people involved, monitoring data and a reconciliation of units filled, incubated and not incubated (9.47).
5What counts as a failure.
Annex 1 sets the target as zero growth: “any contaminated unit should result in a failed APS”. The site then investigates the most probable root cause, puts corrective measures in place, runs normally at least three successful consecutive repeats, reviews production records since the last successful APS with a risk assessment of those batches, and quarantines product made on the line until the failure is resolved. Production resumes only after successful revalidation (9.46).¹
FDA’s guidance says any contaminated unit should be investigated, organisms identified to species level, and the impact on product made since the last media fill assessed. Modern aseptic operations “should normally yield no media fill contamination”, and its recommended criteria depend on run size.³
| Units filled | Recommended criteria |
|---|---|
| Fewer than 5,000 | No contaminated units. One contaminated unit is cause for revalidation, after investigation. |
| 5,000 to 10,000 | One unit: investigation, including consideration of a repeat media fill. Two units: revalidation, after investigation. |
| More than 10,000 | One unit: investigation. Two units: revalidation, after investigation. |
FDA adds that recurring contaminated units on one line, “regardless of acceptance criteria”, signal an adverse trend that should lead to correction and revalidation.³
6How it runs in Seal.
In Seal, an aseptic process simulation is held against the qualified process configuration, its worst-case rationale and a coverage matrix of lines, shifts, operators, interventions, container formats, durations and hold times maintained across runs. Each run’s interventions, participants, monitoring, units, incubation and inspection link to it, and a line change or a contaminated unit reopens only the coverage it affects. A contaminated unit opens an investigation with the intervention window, people, environment and commercial batches since the last successful evidence already in scope.
Aseptic operator qualification resolves whether a person may enter grade B or perform a grade A intervention from current training, gowning, practical, APS and monitoring evidence and any active restriction. Media fill participation counts through the simulation, line, shift, role and the interventions the person actually performed.
References
- 1EudraLex Volume 4, Annex 1: Manufacture of Sterile Medicinal Products, C(2022) 5938 final, sections 7.4 and 9.32 to 9.49, European Commission, 22 August 2022. https:/
/ health. ec. europa. eu/ system/ files/ 2022- 08/ 20220825_ gmp- an1_ en_ 0. pdf (read 6 November 2026) - 221 CFR 211.113, Control of microbiological contamination, eCFR. https:/
/ www. ecfr. gov/ current/ title- 21/ chapter- I/ subchapter- C/ part- 211/ subpart- F/ section- 211. 113 (read 6 November 2026) - 3Guidance for Industry: Sterile Drug Products Produced by Aseptic Processing, Current Good Manufacturing Practice, section IX.A, FDA, September 2004. https:/
/ www. fda. gov/ media/ 71026/ download (read 6 November 2026) - 4PI 058-1, Interpretation of the most important changes in the Annex 1 on the manufacture of sterile products (2022), section 4.9, PIC/S, 28 October 2026. https:/
/ picscheme. org/ docview/ 11495 (read 6 November 2026) - 5Questions and Answers on Current Good Manufacturing Practice Requirements: Production and Process Controls, questions 3 and 10, FDA, content current as of 5 May 2026. https:/
/ www. fda. gov/ drugs/ guidances- drugs/ questions- and- answers- current- good- manufacturing- practice- regulations- production- and- process (read 6 November 2026)
AQuestions and answers
How many units should a media fill have?
Enough to simulate every activity in the routine process. Both Annex 1 and FDA give 5,000 to 10,000 units as the usual starting range. For batches under 5,000 units, the media fill should be at least as large as the production batch, and manually intensive lines need more.
How long are media fill units incubated?
FDA recommends at least 14 days at a temperature between 20 and 35 °C, held within 2.5 °C of target. If two temperatures are used, at least 7 days at each, starting with the lower. Annex 1 asks that the conditions and duration be scientifically justified and validated.
Can a site resume production after a failed media fill before the repeats finish?
Under Annex 1, production on the line resumes only after successful revalidation, normally at least three consecutive successful repeat runs. The PIC/S, EMA and WHO interpretation adds that at-risk batches can be made and certified only once the APS has succeeded and all deviations are investigated and closed.
What if media fills keep failing with no obvious cause?
Look at the medium itself. FDA describes a firm whose repeated failures came from Acholeplasma laidlawii, a mycoplasma in the tryptic soy broth powder, which passed through the 0.2 micron filter used to prepare it. FDA does not expect firms to filter media through 0.1 micron filters routinely; that firm planned to move to sterile, irradiated medium.
Can production staff inspect media fill units?
Yes, if they are trained and qualified to tell contaminated units from clean ones. The PIC/S interpretation of Annex 1 asks that the people whose process is being validated are kept separate from those validating it, for example by having production staff independent of the run inspect the units under QA or QC supervision. FDA asks for QC oversight when QC staff do not do the inspection.