A risk-based approach to microbial control in pharmaceutical production was a recurrent theme at the 21st Pharmig microbiology conference. Susan Birks reports on recent changes to standards
As pharmaceutical product recalls continue to make headlines, Pharmig’s annual event in Oxford, UK, in November 2013 focused on a more proactive approach to preventing microbial contamination. The event drew speakers from around the world, with expertise from the regulatory side as well as the pharmaceutical industry’s top global companies. The meeting covered all the hot topics in microbiology and gave updates on regulation and best practice in contamination monitoring, prevention, identification and risk assessment.
Di Morris, the keynote speaker who was for many years an MHRA inspector but is now with Pharma Solutions, gave advice on the investigation of anomalies known as out-of-specification (OOS) events. She covered many aspects of the procedures that need to be followed when OOS events occur and emphasised the need to monitor and trend data to show that the manufacturing system is properly under control. ‘There is no such thing as normal flora,’ said Morris. ‘You need to know where it is from, how did it get there and how to control it.
‘When an OOS event arises the microbiologist has to become a CSI agent,’ she said, and any investigation must include the nature of the failure, which organisms were recovered and whether they have been seen anywhere else.
All issues should be documented, as even minor events can become significant
A key factor, from a regulatory point of view, is how well the formal review is documented. She suggested that all issues should be documented, as even minor events can become significant. It is also important to document all decisions made and actions taken when an OOS event occurs, as it will show control of the situation and will reassure regulatory inspectors that issues are being handled correctly, however minor they appear.
Not all OOS investigations will come to a conclusive answer but they should not be seen as a failure: ‘You may not be able to find the root cause but you may find other corrective and preventive actions (CAPAs) that need to be fixed, and this is what the regulators would like to see. Therefore write down each factor, even if discounted, as the cause may still require a CAPA.’
Investigations should include a review of the media used and its preparation; the sampling process; where the sample dishes are stored and where the media is kept. It is also important to ensure that there is not a contamination issue in the lab and all lab equipment is clean and validated.
Another suggestion from Morris was to keep a photo library of gram stains to help quickly distinguish any that have not previously been seen. At the end of the investigation, the batch release or batch rejection should be supported by a rationale, she said, adding: ‘Rejection should not be a default position in the absence of any rationale.’ Finally, even if something is ‘within spec’ but the trend is towards failure, action needs to be taken, as ‘out of spec is a step beyond where you want to be,’ she warned.
Andy Matthews, Operations Director at Quantum Pharmaceuticals, looked at effective team investigations. The priority was to assemble a good team to investigate any OOS event. ‘It often needs someone who can have a "helicopter view",’ he said.
He too advised that the aim should be to identify risks before they turn into an event or deviation and he urged quality departments to make the transition from risk management to risk prevention. The microbiologist has a rather difficult role, he admitted: ‘Humans can only see what is there, but a microbiologist’s role is to see what is not there.’
He emphasised the important role that cameras can play both in training and in recording what takes place within a cleanroom. Cameras are under-used in the pharmaceutical industry, he suggested, adding that inspectors like to see process simulations on video.
Any and all micro-organisms present in non-sterile dosage forms may be considered “objectionable” unless demonstrated otherwise
Stephen McGrath, Site Microbiologist at Teva Pharmaceuticals, Ireland, gave an informative presentation on what might be considered ‘objectionable’ organisms from a regulator’s point of view, when manufacturing non-sterile products. Producers of non-sterile pharmaceutical products are required to demonstrate that micro-organisms present in products are safe to release to market and ‘any and all micro-organisms present in non-sterile dosage forms may be considered “objectionable” unless demonstrated otherwise’, he said.
So what is objectionable? ‘It is difficult for those from other disciplines, such as chemistry, to understand that there isn’t just a list of objectionable organisms,’ he said. ‘It is not all black and white in microbiology; pathogeniticy can depend on the interaction with immune systems and the interplay of other micro flora.’
McGrath stated that out of 642 microbiologically-related recalls between 2004–2011, 142 related to non-sterile products and 70% cited the presence of ‘objectionable’ micro-organisms, while 34% of these referred to the presence of Burkholderia cepacia. McGrath quoted the FDA as saying ‘Burkholderia cepacia complex (Bcc) organisms pose a clear and present danger to patient health and safety’, and he suggested now is the time to remove Bcc from manufacturing areas and products.
Similarly, Bacillus cereus, which is associated with foodborne illness from enterotoxin production, was the cause of 50 recalls from 2004–2012. All were associated with alcohol wipes, he said, adding that there was increasing recognition of Bacillus cereus’ pathogenicity and routes of infection.
The USP draft of newly proposed chapter <1115> Bioburden Control of Nonsterile Drug Substances and Products outlines a risk-based approach to the control of potential contamination in non-sterile product manufacturing. According to McGrath, this means microbiologists should be proactively assessing the product and the process, identifying risks and control measures, generating a list of objectionables, assessing the robustness of products, process and controls, identifying potential threat organisms, assessing the suitability of routine test methods, and considering developing specific tests.
He suggested that a proactive microbiological risk assessment should encompass not just the manufacturing process but also raw materials at source, and even the product distribution chain.
Figure 1. The tools are already there to help carry out a risk assessment on objectionable organisms.
Slide courtesy of Stephen McGrath, Teva Pharmaceuticals Ireland
This was a theme also discussed by Paul Newby, Manager, Shared Project Technical Services, Pharmaceutical Microbiology, GlaxoSmithKline (GSK) in his talk on bioburden control of nonsterile drug products. The European Medicines Agency (EMA) had just given approval to GSK’s new inhaled product Relvar. The company had won approval in the US first, and Newby said that with this product, it had anticipated the requirements for a risk-based approach to bioburden control presented in the draft USP <115>. He said GSK had taken a big step forward in showing that it was in control of the system, instead of relying on quality tests at the end of the process.
Newby also noted that Chapter <1115> is unusual in that it looks at all aspects of the product cycle – including product development. It is time, he said, to break down the barriers that split the disciplines of r&d and manufacture because it is important to get quality right from the early development stage.
GSK had focused on ensuring input materials were in control, i.e. that they matched the specs. The company had a range of input controls, such as endotoxin testing, which he said was an excellent quality indicator. The team had looked at the bioburden of raw materials and amassed a lot of past environmental monitoring data to show that they were in control. They also proactively look for specified organisms, including fungi.
Dr Tim Sandle, Site Microbiologist, BPL shed further light on the importance of fungi in his comprehensive review of fungal contamination in pharmaceutical products. FDA figures showed that the percentage of products being recalled due to fungal contamination had risen from 5% between 1990–1999 to 21% between 2000–2010, making it the second highest microbial-related reason for recalls.
Around 50% of fungi could potentially cause infections in humans, said Sandle, and the risks can be from various product forms: inhalers (through inhalation of spores); and creams and ointments that are rubbed into the skin (since many fungi live off keratin, a protein that makes up skin, hair and nails). Such risks increase for those with a weakened immune system or who take antibiotics, he said.
Around 50% of fungi could potentially cause infections in humans
Sandle looked at various sources of fungi as well as manufacturing and cleanroom practices that can help or hinder them. He also looked at current regulatory guidance on fungi, how to identify them and recent studies on the effectiveness of disinfectants.
There is a lack of clarity over the most common fungal genus, he said, and their identification is difficult, yet important when linked to product risk. He advised that checks are made to ensure that disinfectants used are working, as longer contact times may be required.
Barbara Gerten, R&D Manager and Head of Media Development for Merck Millipore, looked at the background to USP chapter <1117> Microbiological Best Laboratory Practices and the practical information in the ISO standards that supplement it. She covered may aspects, such as media preparation, storage and shelf life testing; physical, chemical and microbiological quality control of media; the preparation and maintenance of microbiological cultures; and sample handling, including water samples.
Gerten looked at ISO 11133 Microbiology of food, animal feed and water – Preparation, production, storage and performance testing of culture media, which has recently been revised and is due to be published early in 2014.
One of the useful items of ISO 11133, she said, is a detailed chapter of general terms and definitions, such as batch of culture medium (for which the USP <1117> gives no definition). Also, where USP <1117> contains a detailed chapter on Media Preparation, ISO 11133 gives additional information on documentation from manufacturer or producer, delivery acceptance of products; weighing, water and autoclaving; measurement and adjustment of pH; preparation of supplements (e.g. antibiotic supplements) added after autoclaving; and holding time of molten agar.
Another topic covered was the testing of culture media used for membrane filtration; Gerten said there is a short description in ISO 11133 but the whole procedure is described in ISO 7704 under ‘evaluation of membrane filters used for microbiological analysis’. This standard is currently under revision, she said, with the aim of implementing a more precise description of the method, interpretation and the statistical aspects of the testing.
Finally, Gerten also looked at the new system for microbial reference strains, which uses World Data Center for Micro-organisms (WDCM) numbering. For example, each strain of Staph. aureus has a different reference number in each country or region (ATCC 6358, CCM 4516, CCTM 2103...) but it can now be listed under one WDCM number (00032).
Only a small percentage of labs have so far implemented RMM due to a few myths circulating in the industry
Michael Miller, President, Microbiology Consultants (and owner of rapidmicromethods.com), covered some of the ‘myths and misconceptions’ around the use of rapid microbiological methods (RMM). Only a small percentage of labs have so far implemented RMM and Miller believes this is due to a few myths circulating in the industry, such as ‘RMM are not supported by the regulators’ or ‘they will never replace pharmacopeia tests’ or ‘they don’t support Process Analytical Technology (PAT) or Quality by Design (QbD).’
He reviewed the impact of RMM on industry and gave examples of regulatory acceptance, including both FDA and EMA perspectives. He also went on to discuss how RMM can support QbD and PAT. He suggested that in the future RMM would be ‘the expectation’ rather than a ‘nice to have’.
Dr John Hutcheson, Director, Hutch20, looked at biofilms in water systems from a practical perspective. Biofilms grow very quickly and many disinfection techniques, including UV, may fail to penetrate beyond the outer surface, leaving viable microbes intact. Hutcheson looked at the risks to water quality, how to control or remove biofilms, and he discussed whether biofilms can actually be measured.
Dr Simon Richards, a NSF-DBA Medical Devices Consultant and MD of MEDDEV QA, looked at the microbiologist’s role in medical device production and contrasted the issues that medical device producers face with those of the pharma sector. He outlined the many standards relating to medical devices, including those with a risk management focus.
Edel Fitzmaurice, MD of Fitzmaurice Scientific, and Julie Roberts, Quality Director, J. Roberts Associates, gave a joint presentation entitled ‘Validation, verification and qualification and the suitability of microbiological methods: what’s in a name?’ They looked at these often interchangeable and incorrectly used terms and reviewed definitions from the European and US Pharmacopoeia with a view to distinguishing when to validate and when to verify, and included relevant case studies.
The final speaker, Dr Anthony Hilton, Head of Biology & Biomedical Science, Aston University, took a lighthearted look at ‘Challenging the Five Second Rule’. For those unfamiliar with this urban myth, he explained the rule: ‘You’re hungry, and in the rush to slip freshly buttered toast into your mouth, it slips from your fingers and lands on the floor. What do you do? The rule suggests that if you pick it up within five seconds and eat it, you will be OK.’
As a learning exercise, Hilton set his undergraduate researchers the challenge of designing a series of experiments with the aim of finding out whether how long you leave it on the floor does make any difference. The result was that time, type of floor surface and product all affected the microbial count, e.g. a sticky sweet dropped on a tiled or wood surface had far more microburden than a dry biscuit dropped on a carpeted surface. So the urban myth has some truth!