Myths of pharmaceutical microbiology

Editor's Notes

• #2: Introduction
• #3: Difficult to select myths. Time available, topics selected are: Colony Forming Units – what are they? Microbiology laboratory cabinets – always work? Media growth promotion – can it be skipped? Microbial distribution in cleanrooms – are they free floating? Environmental monitoring parameters – can they be pre-set? Bunsen burners needed to create aseptic space– or not? Identification results– always believable?
• #4: A myth is something that can be widely accepted, but which doesn’t stand up to scrutiny or collected facts. As pharmaceutical microbiology has advanced, some things that have been commonly held as ‘correct’ aren’t necessarily so…or at least not as certain as previously thought. With these myths, some things will be a surprise to some of you; to others you’ll know everything…but hopefully they’re interesting areas to re-visit.
• #5: First – colony forming units – the basis of pharmaceutical microbiology for decades, although now challenged by some rapid methods.
• #6: Should we assume the colony forming unit represents all of the microorganisms in a sample? Or those that could be recovered? No, because of several reasons - Our methods, especially traditional culture based ones, are variable Many microorganisms will not grow on standard media – either ‘viable but nonculturable’ (‘active but non-culturable’) or they are too stressed or the media or incubation parameters are not suitable. Depending on the test method…It is easy to over-dilute, leading to under estimation It is easy to have too many colonies on a plate, leading to confluent growth or over crowding – good paper by Scott Sutton – want an optimal countable range of 25 – 250 CFU per plate We can all make counting or calculation errors I’ll talk about media type, incubation time and temperature later
• #7: Also, a CFU should not be thought of as a single bacterium or fungus – it is a colony forming unit. The ‘unit’ could be made up from one cell or many. There are several situations where this can arise. Example: Poor mixing of a sample before plating out, where cells stick together or become bound to the sample. Bacillus species, for example, are notorious for clumping; This can also be the result of poor mixing of agar plates; Also, with environmental monitoring, if a skin flake lands on a settle plate, this is often carrying more than one organism….I’ll come back to this later.
• #8: Let’s throw in another ‘myth’. Each visible colony on a plate is composed of around one million cells but the organisms within a colony are not all in the same state. Pure colonies come from a single ancestor, but progeny cells are different based on the location. Cells near the middle of the colony can be starved of nutrients and affected by toxic wastes. These cells may not grow when subcultured. Also, central cells are in the stationary phase and will take longer to grow when subcultured. BUT cells at the edge are in the log phase, and should grow faster. Mutations can occur, especially with cells adjacent to each other, leading to genetic diversity – can be an issue with some identification methods...also can lead to transfer of genetic material and antimicrobial resistance. Improvement in obtaining pure, healthy cultures from quadrant plate technique.
• #9: Next myth – clean air devices in labs– are they always contamination free?
• #10: First off – do clean air cabinets always have unidirectional airflow? Well, only when they’re empty. Materials and equipment disrupt unidrectional flow and cause the air to swirl, which can actually spread contamination across surfaces. To avoid contamination, clutter must be minimized
• #11: Here are some illustrations of things that can disrupt unidirectional airflow devices: Objects, Bunsen flames, Broken filter faces, Other obstructions etc. It is important to assess the working area and to check the air velocity is within range where key aseptic operations are undertaken.
• #12: Another clean air device in a lab could be an isolator, as might be used for sterility testing. Although isolators present a barrier, all isolators, contrary to some opinion, leak. What matters is by how much do they leak. This should be assessed before each decontamination cycle using a pressure decay test. This is expressed as pressure drop over time. The size of the leak, if excessive, needs investigation. Here the location of the leak is as much indicative of the contamination risk as the size. A common risk area is with gloves, especially around the cuffs or with pin prick holes. Many users assess gloves post-use by water intrusion.
• #13: Next myth – do we really need to carry out media growth promotion?
• #14: At one level, why do it? Arguments against are: The vendor does it, using type cultures for consistency: Here media growth properties are demonstrated from a low level challenges. And media growth performance can be assessed against a previously released lot, so we can compare growth rates and patterns.
• #15: There could be a case to reducing testing, but here it is important to: Verify the supplier, Ensure the vendor tests the media at all of the temperatures of use that you will use it at, Check that all representative control strains are used and if you are concerned with a particular objectionable microorganism, that this is included, Ensure transportation issues have been assessed e.g. heat shock if the freight lorry breaks down on a hot day. Personally I think confirmatory testing by the receiving laboratory is important to address these factors. Also, if environmental isolates are necessary for inclusion, this can only really be done by the user. There’s no time for the environmental isolate debate.
• #16: Next myth, which relates a little to the colony forming units, is about how microorganisms are distributed in cleanrooms.
• #17: Here it is rare for microorganisms in cleanrooms to be ‘free floating’ Most microbes are found on particles, like skin flakes or dust. Work by Bill Whyte suggests 4 bacteria are typically found on one skin cell, and these are typically around 12 microns in size. This is perhaps an argument for looking for larger particles in cleanrooms. Also, the risk of larger particles falling put of the air through gravity or air striking an object supports the use of settle plates (in addition to air samplers) - if settle plates are in the correct locations. Smoke studies help to decide this.
• #18: The biggest risk from air is mainly that it acts as a vector for contamination and this is a concern in a poorly designed cleanroom. Because microbes do not grow in air and many will eventually die in cleanrooms, unless they are endospores, then where air ends up and whether particles will fall out is important. Airflow design is key to contamination control. With air particles, many microbes will survive for longer on water droplets compared with dust, which means places like wash bays need to be controlled.
• #19: Next myth is about environmental monitoring and that there are universal incubation conditions that we can all follow.
• #20: Actually there are no universal conditions for environmental monitoring because: Not all microorganisms can be cultured. Those that can grow on one culture medium will not grow on all culture media Some organisms can grow, in theory, but will not grow under the physiological state found, or they might grow slowly. The limitations of culture media have been emphasised recently by knowldged the human microbiome of the skin. Much more is found on the skin surface than we recover (see Tony Cundell). The methods we use are limited in terms of accuracy and variability.
• #21: In doing so we need to make some decisions: Will we use one general culture medium –like TSA – and incubate it across a suitable temperature range or use a dual incubation step? Or will we use two media, including one designed to detect fungi and use separate temperature ranges? Do we need to enhance this with selective media if we are concerned about an objectionable microorganism or have a concern about a particular type e.g. anaerobes where nitrogen comes into contact with product. Once we’ve unravelled these, we need to decide on incubation times. This is a huge debate and there is no time to explore it here. I have provided some references on the slide which might be useful.
• #22: How much does this matter? Well, we have to accept the limitations – we can’t capture everything We should run some studies to know we are close to optimal recovery. But most importantly, we need to be consistent. We can do this with our: Locations of monitoring, Frequencies of monitoring, Times of monitoring, Cleanroom conditions for monitoring, And using data to look for trends.
• #23: Next myth – I’ll look at how useful are Bunsen burners for aseptic testing. I’m not referring to loop or slide flaming here; more to the so-called ring of protection.
• #24: The question here is whether flaming flasks for fluid transfer or having Bunsen burners on when carrying out bioburden testing, is necessary? Although it remains common for many labs to use Bunsens, a paper issued back in 1972 argued that it is best not to "flame the mouth of the flask" when transferring fluids, or when pouring autoclaved media into Petri dishes. This is because: This increases the risk of generation of aerosols and create air currents for contamination transfer So a better technique might be: Rapid transfer, Holding the flask or tube horizontal to avoid dust settling.; Use single-use sterile disposable items; Some cases testing in UDAFs.
• #25: Final myth –should we always trust the results from microbial identification systems?
• #26: With identifications there are a number of things that can go wrong. Take the Gram-stain. Errors include: Easy to get a mixed colony, Old colonies lean towards Gram-positives, Over decolorisation can occur through poor technique, Bacillus species can appear Gram-negative. Things can go awry with automated systems: Phenotypic systems are, by their name, affected by phenotypic changes, All systems are only as good as their databases, Cross-contamination can occur.
• #27: With the actual result, the results obtained might be correct under the conditions of the test but is it the result reflective of the organism found? Always question the result of the identification Ask yourself if it is expected from the sample source? E.g. a Gram-negative rod from an aseptic filling area Have I really got Bacillus anthracis? Remember - phenotypic identification systems work on the basis of matching and probability – what you have is the best ‘guess’
• #28: OK – what we have looked at in this presentation: Colony Forming Units – more than meets the eye Microbiology laboratory cabinets – must not be obscured Media growth promotion – probably best to do something Microbial distribution in cleanrooms – risk is with larger size particles Environmental monitoring parameters – need to accept limitations and be consistent. Bunsen burners– probably not necessary Identification results– need to do a reality check on each isolate
• #29: Thank you!!!