Scenario - CFUs present during filling
“The Environmental Monitoring team has recorded an OOS result of 5CFU for an EM sample during vial filling of a liquid batch.”
(The assessor stares into your eyes expectedly)
First thoughts
Loss of EM control. 5CFU in a filling process (i.e. High risk) area indicates a significant loss of contamination control. Potential high risk to batch release unless situation changes with additional information.
Patient risk is three fold:
Presence of bioburden in what should be a sterile or low bioburden product - patient harm up to and including death
Some isolates will bring an additional pyrogen/endotoxin risk
Bioburden in excess (even if the formulation is non sterile) can alter product properties if the isolates metabolise the contents. May also impact stability.
Orientation.
First lets establish the scenario a little more to. Some key questions will influence which direction the scenario will take. Things are a little vague with just the information above.
Is this a sterile filled aseptic product or subject to terminal sterilisation?
If the batch is aseptically filled without further sterilisation - this batch is challenged to ever see the market unless there are mitigating circumstances
If the batch is terminally sterilised the relative risk will drop significantly - but not to zero. Microorganisms can damage product in the time delay before you kill them. Technically in most cases this would not exceed an alert limit for terminally sterilised products (see below)
What type of sample has 5CFU present?
Recovery on an operatives hands. Generally indicates poor aseptic practise or poor disinfection practice - they key question to follow is what the operative was doing prior to the sample being taken.
Recovery from a settle plate or active air sample. Contamination has been introduced somewhere above or around the sample point - again - what innervations have been performed around this time.
Recovery from a surface swab or sample. Generally indicates poor cleaning and disinfection of the surface sampled.
Where (physical location) and when (in relation to the batch filling process) have the CFUs been recovered.
Product contact parts (pipework connections filling needles and stoppering system) present the highest proximity to the product - if you find it here you have to assume it’s in the product
Locations either upstream, around the point of fill and prior to stoppering represent a high level of risk. If the sample is from a settle plate or active air sample you will have to assume it is in the product.
Locations downstream of the stoppering process and prior to capping and crimping have a lower level of risk as the product container is notionally closed (but not fully sealed) but this relies on robust controls for circumstances such as missing stoppers.
Locations within the enclosure but post stoppering and capping are actually quite low risk provided that the remaining upstream EM results are within specification as the product is protected by the container closure system at this point. Controls for checking that all caps have been applied correctly will require reviewed.
When a sample was taken may help indicate its source but in general the earlier on into a batch fill the sample was taken the more risk that contamination has been present in the area for longer - increasing the chance that it may be in a product vial. What can also happen is that samples throughout a long fill are all within spec and contamination is only seen at the end - this may be due to operatives not taking care when disassembling a filling machine or collecting EM plates at the end of a long shift.
Compliance aspect.
The key regulatory context here stems from Annex 1 of UK/EU GMP. Section 9.30 (page 46) lists action limits for viable particle contamination as below:
As you can see - for a terminally sterilised product filled in a grade C (or grade D for some veterinary products) cleanroom - this would not be considered a significant level of bioburden. However if the CFU level recovered is above the alert limit you have set for the facility then you may want to follow the investigation below regardless to understand and eliminate the source.
In general, products subject to terminal sterilisation will be filled in a grade C room as it is substantially easier and cheaper to operate, however formulations that are high in protein or sugars (which encourage the fast proliferation of bioburden) and/or may be difficult to filter could require manufacture in a grade B or grade A area to help manage bioburden prior to sterilisation (but this is rare).
What do you need to see happen?
Depending on your QMS construction this may be recorded as an OOS, an EM Non Conformance or a Deviation. In each record you would expect a level of QA oversight and approval.
The investigation will require input from two key departments. Environmental Microbiology / Environmental monitoring and production.
Microbiology is a specialist subject, and your subject matter experts should be able to provide the following as part of the investigation:
Identification of the isolate to a species level. Isolates that can form endospores or shed endotoxins will increase risk to the product and patient as they present a more persistent impact (endotoxins and spores are significantly harder to kill than bacteria).
An assessment of the ability of the isolate to survive within the product. You generally cannot use the product formulations in hospitability to bacterial growth (if it has good preservatives for example) as a sole risk control mechanism but it is useful information in exceptional circumstances (see below).
An assessment of isolate origin - this typically requires review of the batch filling record and the notes of any cleanroom monitoring or cleanroom microbiologist present to determine the activity ongoing during the sampling period.
An assessment of the controls in place that could stop the isolate entering the product. Were airspeeds and pressure cascades in compliance?
A review of EM trends for the filling area or operative in question. It is important to remember that EM monitoring techniques will only recover a fraction of total contamination within the area and only longitudinal monitoring over an extensive timer period will indicate the level of environmental control that is being achieved within the enclosure. A developing trend can indicate a systematic problem with the filling area itself and may require HEPA testing or smoke studies to understand if airflows have changed.
A review of training for the aseptic operatives and the personal conducting the sampling. Inexperience in sampling can easily contaminate plates and swabs during collection.
Risk assessment and Batch Disposition
The primary risk to the batch is related to the probability that contamination has been introduced where it could enter product vials or contaminate product contact parts. As mentioned above, for open necked vials the only controls in place will be uni-directional air flow within a filling enclosure (UDAF) and aseptic technique. If you recover isolates near to the point of fill or open vials you will generally assume the worst. A 5CFU excursion in particular indicates a significant bioburden that hasn’t been sterilised or disinfected off whatever introduced it to the area.
As mentioned above, for open necked vials the only controls in place will be uni-directional air flow within a filling enclosure (UDAF) and aseptic technique. If you recover isolates near to the point of fill or open vials you will generally assume the worst.
Isolates that have significant physical separation from high risk areas within the enclosure will provide more scope for assignment of a lower risk but this is often dependant on the bacteria properties - spore formers or fast growing bacterial will push the risk factor up again to a reject level.
In most circumstances this is a reject scenario if the product is not terminally sterilised. As ever an exception may come into play if the product is needed urgently to supply high risk patients and there is limited potential to find or make an alternative batch. In that case a consultation with the marketing authority is generally advisable.
If it is terminally sterilised this is generally a low risk deviation unless the product is likely to permit significant microbial growth before sterilisation. Again this will be case by case based on isolate ID and product characteristics.
Note: I have not once mentioned the sterility test. It will provide near zero detection of the sterility failure that CFUs introduced into the filling enclosure unless, possibly you have smeared a massive glob of bacteria directly on the filling needles - and even then I would not trust it.
Finally - the risk assessment should review controls relating to line set up and line clean down - is there a chance that contamination introduced to the area will persist for future batches?
What’s your CAPA?
It is hard to react to a single OOS effectively - especially if the trends of the area demonstrate a positive state of EM control. It is also important to remember an effectiveness check is generally not required - these areas are continuously trended for state of control. If a trend is starting to appear:
Operatives involved may need a period of mentoring or supervision to check that their competency is as expected. This is NOT going and reading a procedure - this is hands on help with your best cleanroom operatives or trainers.
A review of your sporicidal agent rotation may also be warranted
The area may require additional downtime for cleaning if a trend is developing
A developing trend may also require a more intensive test of the areas controls including HEPA integrity testing and airflow visualisation (smoke) studies.
Disclaimer
This document is not a legal document not should it be interpreted in any fashion as a guide. This is a set of notes compiled as part of a personal training program. It is not complete or authoritative. I hope it gives you some value.
This is a personal effort and is not affiliated with my current role or employer.
This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License. If you use it to train someone else, fantastic, but give the QP Notebook a plug. Thanks