Automated Media Fill Inspection

The gold standard for media fill inspection is still manual turbidity checks. Although multiple technologies are available, only a few companies have taken advantage of them. This article provides a summary of the status quo of the process and presents an automated approach using Tunable Diode Laser Absorption Spectroscopy (TDLAS), an already established technology for container closure integrity testing (USP<1207>). Since all microorganisms in question produce carbon dioxide, the growth of the organisms can be detected in the headspace by measuring an increase in carbon dioxide concentration. The growth of microorganisms was demonstrated with glass vials and continuous measurements. Thereby, various growth speeds and different growth behaviors from different microorganisms could be analyzed.

The media fill process is a crucial part in the verification and validation of sterility in aseptic pharmaceutical manufacturing processes. This includes the aseptic processing line, filling equipment, cleanroom environments, and all operator interventions. It demonstrates that the process in place will not introduce any microbial ingress.

For media fills, a sterile growth-promoting nutrient medium is used in place of the actual drug product. This substitution allows growth of microorganisms even in the smallest amount present. By copying the real production conditions as close as possible, a risk to the sterility of the process can easily be detected.

As a replacement medium Tryptic Soy Broth (TSB) is typically used. Equipment start-up, container loading and sterilization, sterilization of all other container parts, filling and sealing equipment, line clearance and all actions by personnel are all tested for their impact on sterility. Typical personnel interactions include tool change, glove adjustments, or any troubleshooting during routine operation runs.

After filling the containers with the growth medium, they are transferred to incubation. During the incubation period the temperature is controlled to a range of 20 °C to 25 °C and 30 °C to 35 °C for a period of 14 days. During this time any nonsterile contamination has sufficient time to grow and change the visual appearance of the growth medium in the containers. This can manifest in turbidity, cloudiness, forming of sediment, or growth of strains. Any evidence of the above will render a media fill as failed. If no change in the visual appearance is observed, the media fill is successful.

Medial fills have a central role in the regulatory environment for validation of the sterility of drug manufacturing and the fill/finish process. Over recent years, regulations have evolved and become stricter since products and manufacturing processes have become more complex. This is best reflected in the new Annex 1 to the EU Good Manufacturing Practice (GMP) guideline which has been in place since
2022.

Media fills, as part of a broader framework for a Contamination Control Strategy (CCS), requires manufacturers to establish a more rigorous, documented, risk-based, and holistic approach to prevent microbial contamination.

• One media fill per shift: To cover all possible working shifts for a new installation, a minimum of 3 consecutive media fills are required.
• In case of significant modification to the equipment, a media fill must be performed.
• If a filling line is decommissioned or inactive for a longer period, a media fill is mandatory.
• A media fill must be performed after the last batch before shutting down.
• The media fill must be repeated twice a year for each aseptic process, each filling line and each shift.
• The number of containers used in a media fill must include an entire batch size, or at least 5'000 to 10'000 units.

The new Annex 1 does not only give more detailed requirements on how and when to perform a media fill but also stresses a holistic approach for process control in aseptic manufacturing.

In this article, the authors will elaborate on the current standard inspection method, the challenges and choice of microorganisms and introduce an alternative deterministic approach to check media fills.

In the context of media fills, the primary concern is the inadvertent introduction of microorganisms that may compromise the sterility of the final product. A wide spectrum of microorganisms can possibly be encountered in cleanroom environments, including bacteria such as Staphylococcus epidermidis, Bacillus spp., Pseudomonas aeruginosa, yeasts (Candida spp.), and molds (Aspergillus spp.). These organisms can originate from personnel, raw materials, utilities (e.g., water or air), or equipment surfaces in general. The identification, classification, and testing procedures for microorganisms relevant to pharmaceutical environments are governed by standards as in the European Pharmacopoeia (Ph. Eur. 5.1.2), the United States Pharmacopeia (US <1116>, <61>, <62>), and the Japanese Pharmacopoeia.

Microorganisms vary widely in their metabolic requirements, including nutrient needs, oxygen tolerance, and temperature preferences. These factors determine their survivability and proliferation. Most of the contaminants in pharmaceutical settings are mesophilic, thriving at temperatures around 20 °C to 40 °C. And they are nonfastidious, meaning they do not require complex nutrients for growth. Media used in aseptic simulations must therefore support growth of these metabolic types, particularly those that may pose a risk under manufacturing conditions.
 

2.1 Aerobic and anaerobic microorganisms

Microorganisms can be classified by their oxygen requirements: 

• Obligate aerobes require oxygen for growth.
• Obligate anaerobes cannot survive in the presence of oxygen.
• Facultative anaerobes can grow with or without oxygen.

In pharmaceutical facilities and during processing aerobic and facultative anaerobic are favored because of the presence of air. Strict anaerobes are not likely to persist in these conditions, as the oxygen in the surrounding air presents a hospitable environment for them. Nevertheless, there are anaerobic microorganisms that are naturally protected in aerobic conditions and can grow.

Media fills are primarily designed to detect aerobic microorganisms, as these represent the most likely contaminants in aseptic operations. Media such as TSB, under standard incubation conditions (20 °C to 25 °C and 30 °C to 35 °C), support the growth of a broad range of aerobic and facultative anaerobic bacteria and fungi. The choice not to specifically target strict anaerobes is both practical and scientifically justified, given their low relevance in environments with active airflow and high oxygen content.

True obligate anaerobes require special environmental conditions for cultivation, including:

• Oxygen-free atmospheres (e.g., anaerobic chambers or jars with gas packs)
• Specific reducing agents in the growth media (e.g., thioglycolate, cysteine)
• Longer incubation times due to slower metabolic rates

To enable the detection of such organisms, the media and environmental conditions must be adjusted. This could involve using anaerobic-specific culture media (e.g., Reinforced Clostridial Medium, Thioglycollate Broth) and controlled incubation in oxygen-free environments. As the same growth medium is used, independent of the detection mechanism, pure anaerobic microorganisms cannot be detected. Media fills, aimed at the detection of organism with an anaerobic metabolism, need a special setup and are only applied in specific cases. However, in practice, such adjustments are seldom made during routine media fills, primarily because strict anaerobic organisms are unlikely to survive or propagate in the highly oxygenated, filtered environments of aseptic pharmaceutical operations.

One relevant example of a microorganism in the context of pharmaceutical manufacturing is Cutibacterium acnes, a non-spore-forming, facultative anaerobic bacillus. It commonly originates on human skin but is protected by a fatty film. C. acnes can tolerate low levels of oxygen – though its growth rate is significantly reduced compared to aerobic counterparts in the presence of atmospheric oxygen [1, 2].
This demonstrates that media fills aimed at the detection of aerobic microorganisms will be able to produce a reliable result for facultative anaerobes but will fail for pure anaerobes (table 1).

Table 1: Direct comparison of aerobic and anaerobic microorganisms.

It is important to point out that the common scenario is aerobic and facultative microorganisms. This means the microorganisms will consume the available O₂ in the headspace and convert it to CO₂. As soon as there is no more oxygen in the headspace the aerobic organism will stop growing.

During media fills, the detection of microbial growth is based on the detection of changes in turbidity, cloudiness, or the formation of colonies. Currently, manual visual assessment is the gold standard for detecting growth in media fill.

The criteria analyzed in the manual inspection of media fills are all directed at the various proliferation of microorganisms. The changes might be very pronounced or very weak depending on the microorganism’s growth rate and preferred growth mechanism. The inspector is responsible for detecting:

• General cloudiness (indicative of planktonic growth)
• Strands or filaments (indicative of molds or filamentous bacteria)
• Color changes
• Sediment formation at the bottom of the container

Personnel assigned to conduct manual turbidity checks must be specifically trained and qualified. Training includes:

• Familiarization with normal and abnormal appearances of sterile media
• Recognition of early signs of contamination
• Practice with reference standards for turbidity
• Regular proficiency testing to confirm visual acuity and decision-making accuracy

This training is essential because visual inspection is subjective. To validate visual inspection competency, personnel is checked for:

• Blind challenge tests using spiked samples with known microbial loads
• Comparison against McFarland turbidity standards, which serve as reference points for estimating microbial density
• Regular requalification, typically annually or per internal Standard Operation Procedures (SOPs)

In addition, all personnel have a periodic examination to test their physical abilities.

3.1 Correlation with McFarland Standards

The McFarland standard is a widely accepted reference to approximate bacterial concentration in a liquid medium. Thereby, the McFarland number corresponds to an amount of bacteria. I.e., a 0.5 McFarland standard corresponds to approximately 1.5×10⁸ CFU/ml and is visually comparable to the onset of turbidity seen in moderately contaminated TSB. These standards are prepared by combining barium chloride
and sulfuric acid to produce a solution with a consistent light-scattering density, mimicking microbial growth. 

While the McFarland scale is not used directly during routine turbidity checks in media fills, it plays a vital role in training inspectors (table 2). In particular, the detection of low-level contamination can be well trained using McFarland standards. The McFarland scale is not used to determine the actual CFU/ml in comparison to real contaminated containers, as it is not designed to do so.

3.2 Limitations of Manual Visual Inspection

Manual visual inspection of media fill containers, performed to detect changes to the filled medium, typically increased turbidity, suffers from several drawbacks:

• It is a non-deterministic method. Manual visual inspection is qualitative and subjective, relying heavily on human perception and experience.
• Human error: Factors such as fatigue, visual strain, and cognitive bias introduce variability into the evaluation process. This increases the risk of false positives or undetected contamination, which can impact product quality and sterility assurance.
• Labor-intensive and resource-dependent
• Accurate visual inspection requires highly trained personnel. Training must include microbiological growth characteristics to be able to distinguish small signs of contamination. However, such inspectors are becoming increasingly difficult to recruit, train, and retain, particularly in regions with workforce shortages.
• Ongoing training and qualification requirements
• Inspectors must undergo routine qualification and retraining, including the use of McFarland turbidity standards for reference, which adds complexity and cost to the inspection program.
• Independent of most container materials: Amber glass and diffuse materials require a refilling of the media fill for manual inspection.

Besides being a tedious procedure, there are aspects to the process of manual inspection that can make it unreliable. The following automated media fill inspection using headspace analysis provides a reliable, trackable alternative which eliminates human error.

Table 2: Example of McFarland standard for the training of manual inspectors on media fill inspections.

Headspace analysis (HSA) using Tunable Diode Laser Absorption Spectroscopy (TDLAS) is already a well-established method for Container Closure Integrity Testing (USP <1207>). TDLAS is based on the absorption of laser light by specific gas molecules at characteristic wavelengths. A tunable diode laser emits infrared light that passes through the gas to be inspected. A detector measures the intensity while tuning the laser over an absorption peak. This allows precise concentration determination for gases such as O₂ and CO₂.

For media fill analysis, detecting the presence or absence of those gases is a direct way to get a quantitative result, making this approach highly sensitive, selective, and quantifiable. Microbial growth leads to a detectable increase in CO₂ and decrease of O₂ levels due to metabolic activity. All aerobic and most anaerobic microorganisms metabolize nutrients in the growth medium, producing CO₂ as a byproduct of cellular respiration. Consequently, a rising CO₂ concentration in the headspace is a strong direct indicator of microbial growth within the vial.

For verification, 7 different microorganisms were evaluated. 3 different sizes of Colony Forming Units (CFU), 10 samples each, were incubated in a 10R glass vial. Headspace volume, media, etc. were chosen according to the regulatory requirements. The CO₂ concentration in the vials was measured daily using a WILCO NEO HSX. Figure 1 and fig. 2 show the growth of the organisms C. albicans and C. acnes. The delicate behavior of C. acnes becomes clearer when observing the growth development. Most of the samples do not show any growth at all. The 2 positive samples were manually identified as such after 11 days, underlying the importance of a 14-day incubation period.

Comparing TDLAS with the manual visual inspection method, the advantages of TDLAS become evident: 

• Deterministic and objective results: It provides measurable, repeatable data that is independent of human interpretation. The method yields precise CO₂ and O₂ concentration values which correlate directly with microbial metabolic activity.
• Elimination of human error: As a fully automated, noninvasive technique, HSA removes the possibility of misinterpretation, ensuring uniform test results across all shifts and operators.
• Data-driven decision making: HSA enables real-time digital recording, trend analysis, and audit-trail generation, supporting data integrity and compliance with 21 CFR Part 11 and Annex 1 expectations for electronic systems.
• Amber glass can be inspected without refilling.
• Reduces training and operational burden: Once installed and validated, HSA systems require minimal operator intervention and do not rely on specialist training, McFarland standards, or subjective evaluations. This significantly reduces the time and cost associated with inspector qualification and requalification.
• Alignment with Annex 1 and CCS: The revised Annex 1 (2022) emphasizes risk-based, robust, and verifiable contamination control practices.

4.1 Challenge of CO₂ Permeation in Plastic Containers 

Plastic containers are commonly used in pharmaceutical and diagnostic manufacturing. They present a distinct challenge due to gas permeability, particularly concerning carbon dioxide (CO₂) – unlike glass containers, which are impermeable to gases. Therefore, plastic containers can have significant implications for reliability of media fill tests, when CO₂ concentration is used as an indicator of microbial growth.

To comprehensively evaluate the extent of CO₂ permeation across a range of plastic container types, a comparative study was designed and executed. The objective was to quantify and compare the rate of CO₂ diffusion in containers composed of different plastic materials and shapes. In total, 10 different plastic containers were selected for analysis, representing a broad spectrum of commonly used formats in sterile manufacturing environments.

Each container was placed in a WILCO bombing chamber to evaluate the permeation rate. The bombing chamber can maintain a controlled and stable environment. The chamber was filled with 100 % CO₂, ensuring that the external environment around the containers maintained maximum saturation. Each container was continuously exposed to this high-CO₂ atmosphere for a duration of 48 hours, during which time CO₂ was able to permeate through the container walls. After removal, CO₂ concentrations in the head space was then measured at regular intervals over a 14-day observation period. The decline of CO₂ concentration within the head space was recorded, as CO₂ gradually permeated into the ambient environment. 

To characterize the dynamics of CO₂ loss, the data were analyzed using Fick’s Law of Diffusion, a fundamental principle describing the rate of transfer of gas molecules through a medium. From this model, the CO₂ flux for each container type was calculated, and the corresponding half-life time determined. The half-life time is defined as the time required for the internal CO₂ concentration to reduce by 50 %.

The results revealed a significant variation in CO₂ permeability among the tested containers. High surface areas and permeability lead to a rapid loss of CO₂ indicating very high flux. In these cases, the internal CO₂ concentration was observed to halve approximately every 1.4 days (fig. 3). In contrast, containers with lower permeability and surface area exhibited a half-life time of 3.6 days (fig. 4).

These findings were combined with the already conducted growth tests in glass vials. Figure 5 illustrates the CO₂ decline with the minimum and maximum permeation of 1.4 and 3.6 days applied to the growth of A. brasiliensis. This simulation underlines the importance of material selection when designing or qualifying containers for use in media fill simulations or any CO₂-sensitive processes. The choice of plastic and container type will have a direct and measurable impact on gas exchange rates, which in turn may influence microbial detection sensitivity.

4.2 Use of Oxygen as an Alternative Tracer Gas in Plastic Containers

Despite its suitability for detecting microbial metabolism in glass containers, the rapid loss of CO₂ in plastics limits its effectiveness in applications such as extended incubation or delayed inspection. For this reason, an alternative approach is required – one that still correlates with microbial activity but offers greater retention in plastic materials. A promising alternative is to monitor oxygen (O₂) levels in the headspace. Due to the low polarity of the O₂ molecule, it hardly permeates through the plastic material, making it a more stable indicator for microbial activity over the duration of a typical media fill incubation period. 

This approach reduces the detectable organisms to the ones with an aerobic metabolism. For most practical cases this is sufficient, since the production environment contains O₂, which suppresses the growth of anaerobic organisms. Hence, they will not be detected with the already established manual visual inspection.

While TDLAS detection of O₂ is generally less sensitive compared to CO₂ (due to weaker absorption lines and lower signal-to-noise ratios), it remains sufficiently accurate for detecting significant changes. In sterile, sealed containers, the head space oxygen concentration begins at approximately 20.9 % (ambient air). A shift towards zero or near-zero O₂ concentration, especially in the absence of other plausible chemicals or physical explanations, can be taken as a strong signal of aerobic or facultative aerobic microbial activity. Even with reduced sensitivity, this change is well within the detection range of modern TDLAS systems and can be reliably quantified.

Although CO₂ remains the gas of choice for microbial detection in impermeable containers such as glass, the combination of O₂ monitoring offers a practical and effective solution for plastic containers. By tracking the depletion of O₂ rather than the accumulation of CO₂, headspace analysis using TDLAS can still provide valuable sterility assurance data in environments where gas permeability is a limiting factor. This approach ensures that media fill simulations using plastic containers remain robust, compliant, and scientifically sound, even under the constraints of material limitations.

Since strict anaerobes require specialized conditions that are not present in cleanroom environments or media fill setups, they are not a primary concern in routine aseptic process validation. This understanding not only supports the use of O₂ as a viable tracer gas in containers exhibiting fast CO₂-permeation but also reinforces the scientific justification for excluding strict anaerobes from media fill design.

The advantages of using TDLAS for media fill analysis, as already stated, far outweigh traditional manual visual inspection – most significantly concerning the elimination of human error, which relies heavily on the inspector’s ability to detect turbidity visually. In addition, attracting and retaining qualified inspectors is becoming increasingly difficult. TDLAS removes these variants and delivers a deterministic and quantitative assessment of the container’s headspace, by measuring the CO₂ and O₂ concentrations with high precision. 

TDLAS provides a substantial improvement in data integrity and traceability. All measurements are digitally recorded, time-stamped, and stored in compliance with 21 CFR Part 11 and EU Annex 1 requirements, ensuring full traceability. This data can be integrated directly into quality management systems, reducing manual documentation errors. Furthermore, TDLAS can easily be integrated into inspection lines. Its integration aligns with current GMP standards and Annex 1’s focus on contamination control, automation, and risk-based validation, helping manufacturers future-proof their aseptic inspection procedures. Although the initial investment in TDLAS equipment may seem considerable, the solution is cost-effective over time.

Dr. Matthias Kahl (Head of R&D at WILCO AG) and Michael Mettraux (R&D Engineer HSA at WILCO AG)

References:

[1] Nagy, E., & Urbán, E. (2001). Antimicrobial susceptibility of Propionibacterium acnes isolates from patients with acne vulgaris. Journal of Antimicrobial Chemotherapy, 47(6), 905–908.
[2] Achermann, Y., Goldstein, E. J. C., Coenye, T., & Shirtliff, M. E. (2014). Propionibacterium acnes: From commensal to opportunistic biofilmassociated implant pathogen. Clinical Microbiology Reviews, 27(3), 419–440. https://doi.org/10.1128/CMR.00092-13

Further references

• European Pharmacopoeia (Ph. Eur.) 5.1.2 – Biological Indicators and media fills. Strasbourg: European Directorate for the Quality of Medicines & HealthCare (EDQM). (For general guidance on aseptic simulation, media composition, and detection endpoints).

• U.S. Pharmacopeia <1116>: Microbiological Control and Monitoring of Aseptic Processing Environments. United States Pharmacopeial Convention. (Discusses environmental monitoring and personnel qualification in aseptic processes).

• Clinical and Laboratory Standards Institute (CLSI). M100 – Performance Standards for Antimicrobial Susceptibility Testing. CLSI, Wayne, PA, USA. (Provides interpretation of visual turbidity and use of McFarland standards in microbiology labs).

• Chesebrough, R. D. (2006). Microbiological Applications: A Laboratory Manual in General Microbiology (9th Ed.). McGraw-Hill. (Chapter on preparation and use of McFarland standards, visual calibration techniques).

• FDA Guidance for Industry: Sterile Drug Products Produced by Aseptic Processing – Current Good Manufacturing Practice. U.S. Food and Drug Administration. (Includes training and visual inspection guidance for aseptic processing personnel).