Annex 1 and microbial monitoring: an interview with Tim Sandle
Microbiological monitoring is one of the most discussed topics of the new Annex 1.
Once again, we spoke to Tim Sandle, microbiologist, author and science journalist, known as one of the leading experts in the field, to discuss in the most appropriate way what challenges the pharmaceutical industry is facing in adapting its monitoring process to the Annex 1 requirements.
Routine monitoring: critical phases
The new Section 9 of draft Annex 1 deals with the monitoring of production processes and classified environments with regard to the counting of viable and non-viable particles.
“Spot” monitoring is no longer acceptable. Frequent (even continuous) monitoring should be considered in the Contamination Control Strategy (CCS). In particular, the focus is on routine monitoring that must be carried out in operations at all critical stages.
What can be considered critical phases?
The reference in the revised Annex 1 refers to sterile products manufacturing. Here each stage of manufacturing should be captured, especially for aseptic processing. This includes line set-up; periods between set-up and the start of filling; during filling; during any lyophilisation; and during oversealing. The product and direct and indirect product components must be protected from contamination at all times.
With how long monitoring should last for, with aseptic processing, then typically settle plates are exposed for the duration of a fill (additional settle plates may need to be used if the fill exceeds the validated plate exposure time). Active air samples can be taken continuously using the new generation of air-samplers that take a sample over a four-period to match the settle plate. For other types of samplers, these should, as a minimum, be taken at the (near) start; (near) middle; and (near) end of the fill.
Finger plates will be taken immediately after connection activity, for any persons present during the fill at a random time during the fill. Plus, after a Grade A zone intervention and after any item of critical equipment is inadvertently touched or where another activity has taken place by which aseptic technique or practice could be compromised.
Surface monitoring (direct contact samples) will take place immediately at the end of the fill. This should not be performed during filling due to the invasive and disruptive nature of the techniques. Contact plates of gowns should be taken from all personnel immediately before they exit the Grade B areas.
In addition, for Grade C and D areas the monitoring should be representative and meaningful, which includes targeting actual activities. For this, a risk framework should be developed whch we can discuss later.
Quality Risk Management: system design, alert limits, action limits and data trends
In the new Annex 1 the QRM principles provide a tool to scientifically identify and assess potential quality risks. In microbiological monitoring, they mention system design, setting action limit and alert levels and reviewing data trends. Can you give us a correct approach to these three activities?
These are three important areas. The first is with the design of the environmental monitoring programme. Here it is important that the programme is based on quality risk management and it takes into account the different types of microbial hazards that may be present and the potential vectors for contamination to end up in the product. Vectors include water, air, surface-to-surface transmission and via people.
Further to design, each organisation should have a documented environmental monitoring programme, and be able to answer the following questions:
• How frequent should the monitoring be?
• How long should the monitoring last for?
• How should the locations for monitoring be selected?
• How should the monitoring programme react to cleaning and disinfection practices?
• Describe the sampling procedure and sample handling
• Describe the sample incubation regime
• Outline the methods for data analysis (statistical data trending)
• Outline the investigative responses to exceeded action levels (and upward trends)
• Describe the execution responsibilities
Alert and action levels are the main way of signalling that something unusual has happened. These can be the numerical values, as described in the Annex. But we also need to look at the frequency of events – especially non-zero counts – and be mindful of the detection of different types of microorganisms.
With setting alert and action levels, for newly built cleanrooms regulatory guidance values can be used. However, after a certain time (such as six months) or when a certain number of samples have been taken (say more than100), each user must set alert and levels appropriate to the facility. This should use some type of statistical technique, such as percentile cut-off or negative exponential distribution. Generally, the percentile cut-off approach is easiest to use and this approach supports data that is not normally distributed.
Interestingly, the revised Annex places particular emphasis upon alert levels as indicators of potential drift, especially of where there are atypical patterns.
The mention of atypical patterns brings us to the third area and trend analysis. I’ve always maintained that environmental monitoring becomes a wasteful and meaningless exercise unless the monitoring has been thoughtfully targeted and the data correctly interpreted to ascertain the ‘true’ trends. Hence, trend analysis is an important aspect of the monitoring programme since data originating from single samples are often not significant.
Data graphs, histograms and statistical process control charts are examples of the tools that can be used and should be applied. Environmental monitoring results which exceed the action level; or where there is an upward trend relating to excursions of the alert level; or where the frequency of incidents exceeds a predetermined cut-off value, represent scenarios which should be investigated.
When putting data together visually it is important to include appropriate information with tables and graphs. This helps to identify patterns and possible reasons for a given trend. Such information includes:
• Identification results;
• Changes to room design;
• Operation of new equipment;
• Shift or personnel changes;
• HVAC problems (e.g. an increase in temperature).
Essentially we need such information to help us to interpret the trend.
Selection of sampling points
Risk assessment can be supportive in one of the most critical issues: the selection of sampling points and related frequency for monitoring. How might his be achieved?
Quality risk management is very useful for determining sampling points and helping to set out the frequency of monitoring. Although there are different risk tools available, for sample site selection my preference is with hazard analysis critical control points (HACCP). This approach is based on understanding and visualising product, people and waste flows.
Through HACCP, the process begins with conducting a hazard analysis, which means identifying what the hazards that can generate a risk are from a process flow. This enables us to determine critical control points, which is about where risk is greatest and seeking to reduce the risk. In areas where risks are higher, the process proceeds to establish critical limits and to establish a system to monitor each critical control point. It is also useful to think about the corrective actions to be taken when each critical control point has an excursion or goes out of control. The HACCP needs to be documented.
In the case of environmental monitoring, we can use a HACCP risk assessment to consider the main areas of risk such as where is personnel interaction is greatest, according to the highest level of activity and perhaps where activities are variable. We need to consider what could lead to contamination entering the airstream and where contact with critical surfaces could occur. This includes understanding if the activity in the area or the item itself contribute to the spread of contamination. As well as the flow of people, materials and waste there may be other indicators of contamination transfer that we especially want to focus on such as airlocks, transfer hatches and locations that are potentially difficult to clean.
An important passage in the new Annex 1 is Section 9.24, in which several sampling methods are suggested, aimed at avoiding interference between continuous monitoring and pharmaceutical operations: Continuous viable air monitoring in grade A (e.g. air sampling or settle plates) should be undertaken for the full duration of critical processing, including equipment (aseptic set-up) assembly and critical processing. A similar approach should be considered for grade B cleanrooms based on the risk of impact on the aseptic processing. The monitoring should be performed in such a way that all interventions, transient events and any system deterioration would be captured and any risk caused by interventions of the monitoring operations is avoided.
With the frequency of monitoring, the use of risk filtering can be useful. For cleanrooms outside of the aseptic core, these can be grouped into different frequency patterns based on the risk presented by different sources and vectors. For instance, we can consider room activity and ask ourselves ‘What is going on?’, ‘When is it taking place?’, ‘What type of equipment is involved?’ and ‘How many personnel are involved?’ We can also considered product exposure risks where we can account for open processing being a higher risk than closed processing, and longer exposure times being at a higher risk than shorter exposure times.
Other factors to consider include room temperature. Is the room cold, warm or ambient? Generally, the risks are lower for cold rooms. Similarly, wet areas will be at a greater risk to dry areas given the greater chance of microbial proliferation in wet areas. We may also want to increase monitoring downstream where there are fewer microbial reduction steps. In essence, we are seeking to monitor where controls are more challenging or where the impact to product is greatest.
Exposed plates: still a valid method?
Another important point in Annex 1: In developing the APS plan, consideration should be given to the following: The method of detection of microbial contamination should be scientifically justified to ensure that contamination is reliably detected.
Grade A: do you consider the use of exposed plates still a valid method or only active sampling should be considered?
Settle plates are an important monitoring tool and they are seeking to achieve something different to active air sampling. With active air, we are assessing the population of organisms within a given volume of air (one cubic metre) within the vicinity of our work area. With settle plates we are looking at the potential for microbial carrying particles to settle out of the air, which can occur due to air current, air slowing down, eddying and so on, and being deposited into the product. Settle plates are particularly useful at working height under unidirectional airflow environments with their locations informed by airflow visualisation patterns. In short, both forms of air sampling are useful, especially within Grade A environments.
What about methods and justification within the CCS? As Annex 1 suggests: Where aseptic operations are performed, microbial monitoring should be frequent using a combination of methods such as settle plates, volumetric air sampling, glove, gown and surface sampling (e.g. swabs and contact plates). The method of sampling used should be justified within the CCS and should be demonstrated not to have a detrimental impact on grade A and B airflow patterns. Cleanroom and equipment surfaces should be monitored at the end of an operation.
New technology can help the continuous monitoring activity: membrane is one of them.
What’s your opinion about it?
With environmental monitoring, environmental control must come first. Without good control, the risks of product contamination are high so we should seek to reduce control weaknesses wherever possible. The range of methods used for monitoring should be based on an understanding of risks and activities, but generally a comprehensive array of methods should be used to assess air, people, and surfaces. These methods are somewhat limited in terms of sample size, the metrology of the methods, and the inability of culture media to grow all organisms (and with many organisms being too stressed to grow). To a degree we can address this by trending, however a new generation of methods offers improved detectability.
It is pleasing to see Annex 1 embrace alternative and rapid microbiological methods and there are a number emerging on the market. This includes technology for bioburden testing, looking for biological markers, and flow cytometry techniques for continuous water system assessments.
With environmental monitoring specifically, there are some emerging systems that can detect colonial growth earlier than is possible with the human eye, scanning for microcolonies using techniques like light excitation. Another interesting area is with spectrophotometric counters. These are particle collection devices that use advances in light scattering, optics and special software, providing real-time data about particles and biological activity in air. With the system, air is passed through a laser and the instrument counts the numbers of particles in a sample of air (inert and biological) through two detectors. Biological particles are detected using a fluorescence detector, looking for three biological markers. These are the metabolites: NADH, riboflavin, and DPA. There are some teething issues with the technology, but it has a powerful potential.
Overall, rapid microbiological methods are to be encouraged. When appropriately selected and validated they provide data that is more: sensitive, accurate and faster, when compared with conventional, growth-based methods.
We would like to conclude this interview with a reflection on the new Annex 1. The Contamination Control Strategy underlying the new annex finds its backbone in Quality Risk Management: assessing risks on a scientific basis and offering a commensurate response to them is the real key to adapting to the new requirements. And like a domino effect, risk analysis has triggered the development of new supporting technologies that are crucial for classified environments.
Dr. Tim Sandle is a pharmaceutical microbiologist, science writer and journalist. He is a chartered biologist and holds a first class honours degree in Applied Biology; a Masters degree in education; and has a doctorate from Keele University.
Interview by Cristina Masciola, AM Instruments – Marketing & Communication Manager