In order to stay within regulatory guidelines, what are some must-have technologies that a pharmaceutical company should implement to meet microbial monitoring requirements?
Tony Cundell, PhD, Principal Consultant, Microbiological Consulting, LLC: Standard equipment found in a typical microbiology lab includes active air samplers, incubators, pH meters, refrigerators, autoclaves, microscopes, Quebec colony counters, laboratory information management systems and microbial identification systems.
Claire Briglia, Technology Specialist, MilliporeSigma: A robust environmental monitoring program is essential in demonstrating that your aseptic processes are in control. Even though it has been around for decades, active air sampling remains one of the best tools for EM. However, in order for air samplers to generate the best data, they need to be in calibration. Is calibrating once or twice a year enough? Having the ability to calibrate your fl eet of air samplers on a routine basis is the best way to show regulatory inspectors that your data is the best that can be. Also incorporating integrated active air samplers into isolators is a must-have. These systems are calibrated in place and can be automatically decontaminated. Lastly media plates are now packaged in gamma irradiated beta bags for isolator applications in which plates are brought in only when you need them.
Paula Peacos, M.S., Senior Consultant. ValSource LLC:
- A good, accurate rapid microbial identification system is paramount. You can’t expect to maintain accurate control if you don’t know what organisms you are dealing with. Some companies are still using very old technologies with limited databases.
- Considering the increased focus on data integrity, it is helpful to increase the amount automation wherever possible. There are new technologies available, the GrowthDirect for example, that will process EM samples and provide the plate counts, thus reducing labor costs and the potential for human error.
- The use of isolator technology as opposed to open filling lines is also becoming an expectation as it is much easier to maintain compliance with <1 CFU requirements for Grade A environments as manual tasks (and operators) are generally removed from the operation.
Dr. David Jones, Director of New Products & Industry Affairs, Rapid Micro Biosystems: The core regulatory guidelines for microbial monitoring have not changed dramatically in recent years, but the guidance documents around these guidelines have become more explicit in calling out the risk of manual observations. The guidelines call for a risk-based approach to contemporaneous secondary verification and have suggested that computer interface and automated technology is one way to reduce the risk associated with human observations.
Dr. Tim Sandle, Head of Microbiology, BPL: Rapid methods will become more important, in order to address the problem with conventional microbiology of reacting several days later to an event. One more recent development with environmental monitoring methodologies is with optical instruments which aim for the real-time counting of microorganisms and non-viable particles from samples of air. Optical spectroscopy is an analytical tool that measures the interactions between light and the material being studied. These instruments work by elastic light scattering.
The instruments measure two things. First, particle counts, where the size of a scattering particle, as it passes through a light beam, is comparable to a certain wavelength of light. The intensity of the scattering is dependent upon the size of the particle. Such systems will detect and quantify particles within a 0.5 to 20 m range. Second, microbial counts. Here a 405 nm laser that intersects the particle beam, so that as a particle passes through the inelastic scattering measures the intrinsic fluorescence of the particle, from the metabolites (such as NADH, dipicolinic acid and riboflavin) inside microorganisms. This distinguishes microorganisms from inert particles.
Poonam Bhende, Assistant Manager, SGS Life Sciences: There are certain technologies and methods for microbial monitoring which are very well-established. Testing should be done at regular intervals. Viable sampling, air and surface bacterial and fungi counts, is a must to observe general trending and determine if cleaning regiments are sufficient. Methods will vary depending on the areas of interest but examples include: air samples, fallout plates, contacts and swabs.
These tools are efficient and cost-effective, and will serve a pharmaceutical company well in meeting microbial monitoring requirements. Looking to the future, we are seeing increasing interest in rapid bioburden and sterility.
There are situations when clients will need to implement continuous monitoring process, for example manufacturers and when validating a clean room. Many of our clients, looking to establish a more robust microbial monitoring program, are requesting that we add microbial identification to their testing package.
Donald Singer, GSK Senior Fellow: With the basis of sound aseptic practices, a clean facility should be monitored to generate data that can be used to measure high risk parameters, i.e. the sanitization and human activity impacts on bioburden in the controlled manufacturing areas during manufacturing. Active air samplers with minimal or no handling of media are must have technologies to accomplish this. Identifications of recovered isolates from low-bioburden controlled areas must be generated by either genetic or other special ‘fingerprints’ technologies. For sterile manufacturing, only by exception will any containment technology besides isolators or restricted access barriers be accepted by regulatory authorities. While we have become accustomed to ‘recovery’ of zero visible colonies on media plates, some new technologies may enhance presence of viable growth not easily detected; any means of developing a better understanding of actual bioburden potential is a step ahead of misinterpreting zero recovery in a clean area with human activity as ‘sterile’.
Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!
If a pharmaceutical company is looking to upgrade its microbial monitoring and testing processes, are there some relatively easy, first steps to take? What do you recommend and why?
Cundell: Contribution Analysis is a management tool that should be applied to pharmaceutical microbiology laboratories to assess their effectiveness. Opportunities to upgrade their processes and increase their impact in pharmaceutical manufacturing include: 1) automation of routine laboratory activities, 2) moving testing from the laboratory to the factory floor, 3) eliminating non-value added activities, 4) leveraging microbial data to improve manufacturing, and 5) the applications of modern microbiological methods.
Repetitive, labor-intensive, error-prone activities in the microbiology laboratory such as plate counting, dilutions for endotoxin assays, Gram staining, data entry, data retrieval, and time-temperature incubation changes can be automated for greater quality and efficiency.
Monitoring that can be moved from the lab to factory floor include active air monitoring, online water monitoring, and in-process bioburden, mycoplasma and bacterial endotoxin monitoring. This would allow for greater in-process control of manufacturing operations and more widespread ownership of the microbial data.
Non-value added activities in the microbiology laboratory that should be targeted for elimination include growth promotion testing of general microbiological growth media, disinfectant efficacy studies, microbial test parameters in product stability studies, microbial monitoring in closed isolator systems, and much unnecessary, duplicated logbook maintenance. Commercially prepared reagents, media and inocula should be used to free up analyst’s time.
Areas where microbial data collection, tracking, and trending could be leveraged include continuous laser-induced fluorescence viable and non-viable particle monitoring in cleanrooms and water systems, computerized tracking and trending for process control and improvement, and increased efficiencies in microbial data deviation investigations.
Briglia: Microbial testing can be very labor intensive and there have been some changes to existing product platforms to improve workflow. For example, the latest upgrade to our Milliflex bioburden testing platform includes hardware that doesn’t require autoclaving and very easy no membrane handling directly to the media plate.
When there are improvements to workflow, the risk of both false positives and false negatives is reduced.
Peacos:
- Any technology that that decreases the amount of direct operator contact with the sample as it greatly reduces the likelihood extraneous contamination. For example, I recommend the use of self-contained viable air monitoring products as the operators never come into direct contact with the agar medium. There are also no parts to sterilize or assemble.
- Any technology that can upload the raw data to a computerized system and perform automated calculations is also helpful as it eliminates the need for transcription, saves time and labor while reducing error. Some laser particle counting systems for example can do this.
- Operators, especially if non-microbiologists are performing the actual sampling, should be trained to understand how microbes move through a facility and the limitations of microbiological methods in terms of detection levels. They should understand that effective contamination control is obtained primarily through exclusion as opposed to removal. Simply training them on how to perform the sampling is not enough as they will often not see an issue developing in the area or recognize an error.
Jones: If a pharmaceutical company is looking to upgrade its microbial monitoring and testing processes the first thing they should do is move to an automated alert system. Too many facilities receive 483s because they didn’t act on results that exceeded their own action and alert specifications. Companies can automate the review of their data by installing an environmental monitoring software package to facilitate rapid and frequent review. Inclusion of automated microbial detection systems and direct communication to the software facilitates better data security as well as better control.
Sandle: It is important to invest in good environmental monitoring software, using a bespoke package rather than trying to configure a standard LIMS system or cope with Excel (which doesn’t stand-up well to data integrity requirements anyway). The level of complexity in relation to many cleanroom operations and the regulatory requirements for root cause, appropriate CAPA and for linking events together as part of a holistic contamination control strategy require increasingly sophisticated analysis and only software that is capable of producing meaningful trend analysis can do this.
Outside of this, anything that helps to automate the process is good, such as a facility monitoring system to connect up particle counters; and where portable counters are required to use digital capture and wireless technology. Having the ability to start active air samplers with time delay functions, or remotely, also helps to reduce the risk of the technician using the devices from cross contaminating them.
Bhende: Within regular sampling events, clients should consider getting more from their environmental monitoring samples. Characterization of the isolates is a start. Depending on how specific a client needs to be, biochemical testing and genetic ID work can be an important part of making sure their areas are clean and safe.
The type of organism identified might suggest the origin of contamination (human, water, process), the level of risk, and how a client may need to modify their cleaning protocol.
Singer: Each company has its own culture that leads the way for it to determine what to consider for microbiological monitoring. A first step can be an evaluation of the current program to see if there are any efficiencies to be gained by reducing monitoring in low risk areas. This can be followed by evaluating monitoring of high risk areas for frequency/method effectiveness and determine if recovery capabilities and identification methods adequately match control outcomes. There is a common approach in many companies to perform extensive monitoring to comply with regulations, yet frequent monitoring of low risk areas or operations often does not provide requisite value. The abilities of microbiologists performing and evaluating results should be a key consideration when upgrading a microbiological monitoring process. More data lead to a need for more analytics, and more identifications lead to a need for better analytics methodology. The most important considerations for any new monitoring process should be: will the change fit the intent of the microbiological monitoring strategy and what infrastructure (human, lab, supervision/oversight) will support the change?
Do you foresee pharmaceutical companies implementing the strict microbial monitoring strategies used for sterile product production on other types of products – such as solid dosage?
Cundell: No, definitely not. In general, I view microbial monitoring in sterile product manufacturing, especially in isolator systems, as excessive and compliance and not risk driven. As discussed in USP <1115> Bioburden Control for Non-sterile Drug Substances and Drug Products, environmental monitoring is not a critical process control in non-sterile drug manufacturing. Periodic microbial monitoring, at a frequency and intensity based on microbial contamination risk, may be useful to confirm that the environment controls, HVAC system, cleaning program, material flow and operator activities are performing as expected, but being periodic and non-critical, is not linked to product release.
Briglia: Non-sterile products that contain a high percentage of water pose a significant risk of gram-negative bacterial contamination; thus, this will continue to be a regulatory focus and will require strict microbial monitoring.
Peacos: I don’t think it’s practical to do this. Control levels should be commensurate with risk to the patient. The key is proper identification, evaluation and mitigation or elimination of those risks. For example, if the patient population is immunocompromised, or there is a known risk with the product type (such as B. cepacia complex and nonsterile liquid preparations) I would expect to see more robust control strategies than I would for other products with less inherent risk. The amount of data is not nearly as important as the quality. It is important to understand what data you need to collect and what that data actually means to the product and process control. Is it providing what you need it to in terms of value?
Jones: Yes, we do see the FDA suggesting more routine environmental monitoring for all types of manufacturing. Frequent environmental testing and monitoring is an important tool for demonstrating that a facility is in control.
Sandle: In answer to this question, the monitoring should always be proportionate to the risk and the risks with solid dosage are different, and they need to consider places and locations where objectionable organisms may occur and focus here. In addition, there should always be a meaningful microbiological environmental monitoring program based on risk to the process, and drawn up using risk management principles. Tools like Hazard Analysis and Critical Control Points (HACCP) are useful for this purpose.
What is in danger of happening is ‘regulatory creep’ towards a higher level of monitoring. If a firm is carrying out a sound, risk-based monitoring program this should be sufficient and industry does need to ensure that what is being asked for is appropriate and proportionate.
Bhende: We are seeing some clients in non-sterile environments stepping up their microbial monitoring strategies. Specifically, clients concerned about B. cepacia. They are adding HEPA curtains around their non-sterile rooms, either as a preventative measure or in response to FDA findings.
Singer: I do not and hope this doesn’t occur. The only similarity across monitoring strategies should be that they are based on facility and operation ‘cleanliness’ intent and finished product microbiological control. In general, control for a solid dose product considers any inherent bioburden and prevention of contaminants, whereas control for a sterile product considers all bioburden as contaminants. The differences stated lead to different strategies which have impact on quality and safety for the patients. Although I hesitate to mention quality and cost together, they are both relevant to access of drugs to our patients. Monitoring of a solid dose operation similar to a sterile dose operation could add substantial cost to the solid dose product to a patient. Monitoring of higher risk non-sterile products that are administered into the lungs or into wounds may be more similar to monitoring of sterile product operations, since risk and cost are a balance taken into account.
What are some best practices a pharmaceutical company should put in place to collect, store and analyze microbial monitoring data?
Cundell: Best practices would include bar-coded label generation, microbial contamination risk assessment, sample location mapping, automated plate reading and data acquisition, application software to manage the environmental monitoring program that interfaces with LIMS, and computerized data analysis and report writing.
Briglia: Many companies have implemented LIMS and electronic notebooks into their facilities. Regulatory inspectors are trained and now expect to see these systems. If you do not have a LIMS, you should be planning to have one very soon. Inspections are so much more efficient if data is easily accessed.
Peacos: I think thorough, in-depth training is paramount. Microbiological test results, particularly visual colony counts, can be subjective at times. Training must be consistent, and it is important that trainers ensure complete understanding. Competency should be demonstrated. It is also important that the training program ensures that operators perform the tasks in the same manner to reduce variability. Training to this level is admittedly time and labor intensive, but the results in terms of data accuracy and integrity are worth the effort in my opinion.
Jones: Moving away from paper is an important first element in collecting, storing and analyzing the large amounts of data that is being produced in manufacturing facilities. The next step is to automate the analysis of this data in a secure database that has a full forensic audit trail. The third best practice is to learn from this data and use it to improve the manufacturing and cleaning processes.
Sandle: The important consideration remains data integrity. The analysis of microbial data is only any good if the data is what it should be. The second important point is when designing data systems is to focus on both the input and output, there’s no point investing in a system, for example, if the right type of trend analysis cannot be performed. With storage it is important for the pharmaceutical firm to ensure that they keep ownership of their data, there are risks with handing over the data to a server in another country for example. What happens if the company providing the software goes bust or the relationship ends?
Many packages come with standard reports, but often customizable reports will be required. The ultimate aim, I think with environmental monitoring systems, is with being able to work out where contamination has come from, how often it is occurring, and where it is going. Lastly, assessment is not only about numbers (like colony forming units); it’s also important to look at microorganism trends and to assess these against benchmark data.
Bhende: As an independent contract laboratory, our clients are interested in receiving the reports and raw data. In certain cases, SGS also provides trending reports for clients on request.
When clients provide us with specifications, our team can alert them to any OOS so they can take the appropriate action. Then SGS can retest once they’ve resampled.
Singer: Best practices today are use of a computerized entry and storage data management system that has acceptable controls in place to ensure the integrity of data. The system can be solely for microbiological data or can be a multi-disciplinary laboratory data entry and storage system. The handling of some microbiological data may initially be manually generated, and consideration for the ALCOA approach to data integrity throughout the data lifecycle is a good for compliance. When possible, use of barcodes or automated reading of information relating to media, site location, sample types, etc., are recognized as being best practice. Analyzing microbiological data can take many forms, whether evaluating trends by graphics or spreadsheets. Utilizing the latter as electronic tools for periodic data evaluation and generation of reports for management is currently best practice.
In the near future do you see pharmaceutical companies moving away from large cleanrooms to processing sterile products in isolators, gloveboxes or RABS? If so, why?
Cundell: Yes, this trend is well underway. It is generally believed that the use of isolators and restricted access barrier systems in new or upgraded aseptic processing facilities reduced capital and operating costs, excluded people from critical aseptic process steps, and improves sterility assurance. Sterility testing isolators are the norm in big pharmaceutical companies.
At the 2003 AASP Aseptic Process Workshop, Jack Lysfjord, who worked for Bosch Packaging Technology prior to becoming an industry consultant, discussed the evolution of the technology from conventional cleanrooms to RABS and isolator systems. He estimated that there were in excess of 2000 filing lines worldwide, with about 10% using barrier technology, and twenty new systems delivered annually. With this rate of introduction it was estimated that it would take a decade or two for all the conventional filling lines at use at that time to be replaced by barrier systems.
In a follow-up 2012 survey with Michael Porter from Merck, they found that the number of companies using barrier isolation went from 32 in 1998 to 131 in 2012 representing an average annual increase of 3.5% or fourfold increase. In terms of the number of filling barrier isolators globally the number increased from 84 in 1998 to 490 in 2012 representing an average annual increase of 14.5% or a six fold increase. If we take a conservatively low estimate of 2000 filling lines, this is a mere 25% installation of barrier isolators. How many isolators have been installed and in use in 2018, you may ask? Based on the historic growth rate, the number should exceed 800. This installation rate is higher than Lysfjord predicted in 2003 but lower than would most beneficial for companies, regulators and patients.
Briglia: FDA has stated several times that all new aseptic processing lines should be in an isolator or RABS. If you construct a conventional clean room instead, your facility will be scrutinized on every inspection.
Peacos: Absolutely. The level of sterility assurance and overall process control in an isolator unit or even a RABS versus an open filling line is just so much better. By reducing/removing the physical presence of the operators the likelihood of contamination is reduced, and patient safety is better assured. This is critical when you consider the limitations of sterility testing and microbiological monitoring methods in general. Furthermore, this technology has been available for some time now, and the units available today practically run themselves. There is really no good excuse not to use it. The use of isolator technology is rapidly becoming the industry standard as most new facilities are incorporating it into the design.
Jones: Yes, in an effort to reduce the risk of human contamination the introduction of robotic manufacture in RABS or other isolation systems will become more common. Many companies are already investigating the options in this field and as these become commercially available companies will transition their processes to these kind of manufacturing environments.
Sandle: This is certainly a trend and a welcome one. Environmental monitoring is always limited due to the small sample sizes and limitations in terms of the meteorology of the instruments, plus little of it is in real-time. Therefore, pharmaceutical manufacturers should be investing in technology that reduces the risk of product contamination and barrier technology provides the means to do this. When operating effectively, barrier technology serves to separate the operator from the product, minimizing the likelihood of operator derived contamination. In addition, devices like isolators and RABS provide unidirectional airflow, of a low particulate count, and they have the capability for the internal environment to be decontamination using automated processes. All this adds up to better control and a lower contamination risk.
Bhende: We are not seeing a huge shift right now. There are certainly situations where clients have specific requests for isolators. However, there are situations where that’s not appropriate or possible, for example with medical devices – the samples are large and wouldn’t fit in an isolator, and must be testing in a cleanroom environment.
Singer: New drugs with smaller patient populations and smaller batch sizes do not need large cleanrooms to incorporate the controlled manufacturing, thus specifically designed isolators or closed RABs may be adequate with relevant activities and environment surrounding them. There will continue to be products for the larger populations, and these may still be produced in cleanrooms with RABs. I can see future cleanrooms being designed as smaller to fit more operations under one roof or more interchangeable operations, but the parameters for control of human operators and materials must continue to be designed to fit an operation for contamination control. Even if we start to design and use smaller areas for clean operations, we should continue to train and build confidence in our operators with ongoing knowledge and experience in the impact human intervention has on the microbiological control of sterile manufacturing. Robust design, validation and training are key aspects of sterility assurance.