The Hottest Topics in Microbiology

Recently I asked participants in a senior level think tank, what were the most challenging microbiological issues on their “to do” lists. The answers included:

• Correlation of field failures for bacteria to hygiene in the manufacturing environment or against efficacy testing.
• People.
• Perception that you can validate manual procedures leads to miscalculations.
• Use of contract workers and high turnover.
• Seems like a lot of us are disinfecting/sanitizing but what do we actually intend to achieve.
• Dealing with the ambiguity of bioburden control involved with microbiologically controlled products and the perception (by some) that we are making sterile products.
• Need to educate about risks.
• Risk assessment of environmental organisms and the need. for better knowledge and control. Use the risk assessment for how significant is the organism for the consumer. How do I ID the organism and how do I do the risk assessment –the regulators are all over the place.
• What truly is an objectionable/specified organism–base on a risk assessment (documented) for YOUR product type and also any USP specified organism.
• Contract manufacturer fights us on requests for objectionable organisms? Our spec is just specified orgs. Reply: USP <1111> requirement to assess any organism you pick up in the environment.
• Raw material suppliers have no idea about the market they are supplying and refuse to do it. There is a price on CMOs and sometimes cheaper and easier to bring the baby back home. You may have to pay for your requirements.
• Non-viable particulate count room qualification per ISO regulations –rooms are controlled and we claim class E –now required “new class E = D” for non-sterile products? Everyone talks about “risk-based” EM and EM trending and yet I have yet to see one example of how anyone is actually DOING it! And have never seen anyone using relevant statistical analysis of EM data.
• I was so excited by<1115> but it says nothing: product specification is NMT 0 cfu/ml.

So…ouch. You can hear the pain and then start to filter out issues. Actually when you sort through the list, there aren’t as many different concerns as it first appears. Here is what I found:

1. Increased emphasis on non-sterile products. Increasing regulatory scrutiny. Unclear requirements, but nevertheless expectations for controlled manufacturing environment, microbial limit testing and preservative efficacy testing if the product is preserved. The only regulatory body who has published guidance for non-sterile environments is the World Health Organization (WHO). Neither the EU GMPs or the US GMPs or FDA guidances address this topic. The EU is currently revising Annex 1 on Sterile Products and in their concept paper on the revision indicated that the revision might address suitable environments for manufacturing non-sterile products. Since the revised Annex has yet to be published, the jury is out.

What should you be doing?

Regarding non-sterile products there are two references which might help.

The first is the USP General Chapter<1115>, Bioburden Control of Non-Sterile Drug Substances and Products. As one of the think tank participants pointed out, it looked promising but is actually quite confusing/confused. On the one hand, it advocates “a pragmatic scientific approach to the management of the microbial bioburden in non‐ sterile drug products in keeping with patient risk and contamination control objectives based on risk management principles”. The chapter is advocating process control rather than reliance on end product testing which is commendable and the desired state.

<1115>is an informational chapter and not an enforceable regulation.

On the other hand, the chapter advocates for any environmental monitoring program to be risk‐based in terms of sample selection and frequency and to be used to confirm microbial control and not linked to product release.

This too is commendable as a statement for those well versed in process control as opposed to product control. However, in practice it often results in a check box exercise which is far removed from true risk management and monitoring. Environmental monitoring is a RISK MONITORING tool and not, as commonly and mistakenly believed, a risk reduction measure. EM samples, represent intelligence gathering, allowing ongoing assessment and understanding of the state of control of the environment. Increased counts, and in non-sterile production, since the baseline counts are generally far higher than in sterile production, it is easier to identify trends earlier and to respond. But if the EM is disconnected from production of specific products and batches, and, is frequently conducted during “at rest” conditions, the purpose of the monitoring is lost, the responses don’t happen and failures in Microbial Limit Tests are then written off as “laboratory error.” According to Tony Cundell, one of the authors of the chapter, the intent was for improved and cost‐ effective bioburden controls for non‐ sterile drug products to reduce the number of product recalls arising as a result of microbial contamination. No question that this necessary – especially after the recent episode of contaminated sunflower seeds resulted in recalls of many brands of health bars including my personal favorite the Cliff bar!

The second essential reference is PDA’s Technical Report #67: Exclusion of Objectionable Organisms from Non-sterile Pharmaceuticals, Medical Devices and Cosmetics. This is a points to consider document that addresses product considerations, risk management, how to identify clinical considerations related to defining objectionable organisms for a particular product as well as process design to mitigate microbial risk – all valid and appropriate considerations in setting up a microbial control strategy.

2. How can we make risk assessments less burdensome and more useful/effective?

Assessments need to be simplified. The risk question needs to be sharpened. For example, for non-sterile products what is the risk question if performing a risk assessment for the manufacturing process? How about “what could go wrong resulting in unacceptably high total count or presence of objectionable organisms in the product” However, from a microbial point of view, the answer is not straightforward. It ties in with the above observation that “Seems like a lot of us are disinfecting / sanitizing but what do we actually intend to achieve.” The answer to that is that we need an environment, populated with operators appropriately gowned to ensure that things do not go wrong resulting in high total counts or objectionable organisms. Since the two are related and since microbes multiply by doubling in short periods of time, the answer to the disinfecting/ sanitizing is the following:

The purpose of disinfecting in a non-sterile area to maintain routinely low levels of microbial contamination with the intention of preventing product processed in the environment from ever exceeding the Microbial Limit specification for total count or objectionable organisms.

3. Environmental monitoring – what is needed and how do we reply risk assessment and contamination recovery rates (per USP <1116>).

The answer to this question is a follow on from #2 above for nonsterile products.

What is needed is a truly risk-based program, with samples at those areas MOST LIKELY to disseminate contamination to the process and therefore to product.

4. Control of outsourced manufacture and testing With respect to microbiological risk – any program we have in our own company needs to be mirrored in some manner at our contractor to provide the necessary assurances. We audit for that and if there are concerns they must be remedied under the CAPA program before signing a contract. Ongoing evaluation of the supplier for bioburden controls and microbial test must include review of EM data.

5. How much work should we be doing with identification of organisms and how does that tie in with cleaning and disinfection programs and field failures, hygiene and the people factor?

There should be an SOP defining identification policy. Firstly, sufficient isolates must be collected to obtain a baseline flora for an area and that is more than only identifying those isolates where the alert or action level was exceeded. Thereafter, the policy should be based on potential significance to product and therefore to patient and periodic across the board identification to see if there are new isolates recovered and to test recovered wild types for sensitivity to the disinfectants used in our sanitization regime. A sudden increase in fungal counts is reason for a hysterical response – yes I did choose that word – rapid and aggressive action to identify and seal off the source (often cracks or building movement when there is heavy digging nearby from construction) and repeated cycles of sporicidal disinfectants. You do not want to be the next facility shut down for fungal contamination (Google “compounding pharmacies and fungal contamination” for more details)!

And beyond that consider the following hot topics with the regulators, all of which have implications within the microbiological area:

• Quality Metrics and Culture
• Data Integrity –the microbiology lab has a special place

in warning letters for “throwing away plates” or “failure to record colonies” all of which is easy to do because there is no trail whatsoever (never mind a part 11 compliant, time and date stamped, automated computerized trail). In the microbiology laboratory, most companies still work with an operator manually examining incubated test samples and recording the outcome. The temptation to cheat is facilitated by the ease with which it can be done – which leads into cultural issues and independent verification checks.

• Multiplicity of regulations (IVT has a link with over SEVENTY microbiological references)
• Revision (planned) of Annex 1 to the EU GMPs (Sterile Products annex) and updated ISO 14644 (in effect but are contractors ready for that?)

And how do all of the above tie in with the recent massive recall of health / granola bars due to listeria contamination of sunflower seeds? To conclude – new issues? Not really. More of a new focus on non-sterile manufacture. Is non-sterile manufacture, as traditionally perceived, less complex than aseptic processing? In fact from a microbiological perspective it might be MORE complex – because the goal is poorly designed. Sterile means the complete absence of viable organisms, or at least bacteria and fungi. What does Non-Sterile mean: NMT 100 cfu / g total, NMT 10 cfu/g fungi and no objectionable organisms. Try telling the bugs to stop replicating at 100 or the fungi at 10. If your microbial control strategy does not strive to keep them out, they will sooner or later take over and you will have OOS Microbial Limit Tests. A control strategy is defined in ICH Q10 as:

“a planned set of controls, derived from current product and process understanding, that assures process performance and product quality” A microbial control strategy is based on risk assessment, implementation of controls directly related to understanding of microbial risks posed by product and process such that the controls assure process performance and product quality. i.e. no MLT failures. Environmental monitoring is one of the risk monitoring tools which allows you to show ongoing control and while not directly related to batch release decisions, is definitely one indicator that the batch might be fit or unfit for release from a microbial perspective.

One final comment. No think tank or article on hot micro topics would be complete without discussing rapid methods. There are many available. For the most part, pharma companies seem to have implemented these mostly in the area of microbial identification. However, there are increasing interesting offerings in the area of real time counts of viable and total particles where the total particles include and can distinguish viable organisms. This is an evolving area and…watch this space!