Media Fill Failure – How to handle!

I initially hesitated before writing this article since media fill is not directly related to emulsions, and I am not an expert in the aseptic process. However, during my production journey, I encountered one of the most challenging times while conducting media fill simulations. When I joined this particular company, it was a large organization but not a multinational one. However, after several mergers and acquisitions, it grew into a large and bureaucratic company, making it difficult to comply with all the procedures that were not in place before.

In the past, our team produced around 16,000 liters of vaccine per week, equivalent to 320.000 units in two presentations. This means we were producing 64.000 liters per month and 768.000 liters of emulsion yearly. It took us about 3 to 4 days to fill the products, including setup and disinfection, and we had to do it for 4 to 7 days a week in 3 shifts. We could produce 150.000.00 doses of vaccine per year, which was all water and oil emulsion. However, we had to stop production for six weeks each year to perform media fills, which was a challenging decision because we would lose the opportunity to fill almost 90.000 liters of product. Nonetheless, we had to ensure compliance and perform media fills three times to validate the process. We also needed to perform the test before and after each intervention in the filling room and every six months.

Media fill is a well-defined online process, so I will not describe it again. Below, I transcribed an FDA definition:

“A “media fill” (sometimes known as a “process simulation”) is the performance of an aseptic manufacturing procedure using a sterile microbiological growth medium in place of the drug solution. Microbiological growth medium is used in place of the drug solution during media fills to test whether the aseptic procedures are adequate to prevent contamination during actual drug production. A media fill is one part of the validation of an aseptic manufacturing process.

The media fill should evaluate the aseptic assembly and operation of the critical (sterile) equipment, qualify the operators and assess their technique, and demonstrate that the environmental controls are adequate to meet the basic requirements necessary to produce a sterile drug by aseptic processing. The media fill does not validate the ability of the filter to sterilize growth media.” – (FDA Media fill)

Basically, you will test your filling process with a rich media and hope nothing will grow there.

After long discussions, we faced difficulties performing with such large volumes and a long filling time (approximately five days). We were instructed to make three full batches of our total volume, but this would be incredibly expensive – potentially more costly than the entire vaccine batch. Fortunately, we found a media fill expert in the company who suggested a theory about the media fill size. Due to our large volume, they explained that more than 10.000 units would be required to simulate the real process. Instead, we needed a bigger volume to cover the 20.000 liters formulation tanks, the buffer tank, the filling machine, and all remaining piping systems. We ultimately agreed on using 4.000 liters of media, with tryptone soya broth (TSB) at around 160.000 units – still a lot, so we were allowed to do 20,000 units of media followed by 120.000 units of water, and then another 20.000 units of media three times to validate our process; in this case, we only used 50% of the filling volume of our 50 ml bottle and thioglycollate Medium was required since we do not use gas to create an oxygen-free environment or anaerobic bugs are isolated from the environment.

Our team started implementing a highly detailed and systematic protocol to ensure our operations’ utmost precision and quality assurance. We pay close attention to every process step, carefully monitoring the individuals working in the filling areas and the tubes being used. We tracked the time taken for each task and closely followed any breaks taken. We aim to minimize the need for any intervention as much as possible since there is no such good thing as one intervention.

We were meticulous during the procedure, treating it like a finished product. However, despite all efforts we received the unfortunate news that we had failed. Typically, one contamination or <0.1% of the tubes are possible when filling a large number of units, but we had multiple contaminations, a significant deviation from the expected outcome.

The team was frustrated, I was frustrated, and the worst was the leadership team was frustrated; it is needless to say that the quality assurance department immediately stopped any process and interdicted the filling line immediately; what a disaster; several phone calls from the sales team, directors, and global positions that we had never heard about it – it was complete chaos for a US$ 250 million dollars per years business.

The situation was quite complex, as the team never faced a problem. The issue was related to the investigation of a media fill failure; no one at the site had any prior experience dealing with it. The pressure was mounting, and there was a sense of panic among the team members, who didn’t know what to do. However, we were fortunate to have good leadership that could control the situation until we started the investigations.

To handle a critical situation, the team calmed down and filtered communication to ensure only necessary information was being shared; also, a task force consisting of their most experienced and skilled members was created, with a similar approach used in the Apollo 13 mission, where the team gathered together in a room and started brainstorming a plan to address the issue. The task force comprised different team members such as MSAT, manufacturer, quality, technical service quality control, and engineering. Each team member was responsible for its function and reporting, but we had one voice, or only one report would be shared with the corporation.

Needless to say, at first, the blame was placed on the media sterilization process; however, this hypothesis was soon ruled out by the results of the positive and negative controls. Thus, the team needed a more elaborate plan to solve the issue.

This first step was to find the contamination source, the type of microorganisms that had contaminated the vials and we had the best microbiology in hand with years of experience and we identified several  Staphylococcus spp. and Penicillium spp as the main contaminant.

Now we know our “enemy,” and we just need to see the terrain we are fighting, or in other words, where it comes from.

The idea now is to generate data and analyze all the large amounts of data we received from multiple sources, such as the engineering and QC, where sample time, air samples, particle counts, interventions, and others were poor in the investigation. The hardest work was filtering all these data, but when you have a suitable protocol, you can track –When + Where + Who the contamination came from, or at least you have good indication. And we had a good protocol. Just by chance, we were asked to participate in an experimental procedure despite heavy protests. The procedure aimed to identify the representative bacterial isolates recovered from cleanroom operators’ garments. To accomplish this, we used 16S rRNA gene sequencing to determine the origin and identity of the bacterial isolates collected from the garment surface. The process involved testing the cleanroom operators’ garments before and after their work shifts and using the direct agar contact method to collect bacteria from the garment surface.

As part of our investigation process, we requested all internal procedures related to clean room work and disinfection in the company. Our primary focus was to identify any potential gaps in the existing procedures. We took a closer look at several aspects, including particle counters, clean room validation, and qualification, operators’ activities, materials used in the area, equipment available, and the timing of sample collection that led to contamination. Furthermore, we analyzed the media fill tapes; we filmed all procedures for days (back then, we had taped because there were no cell phones to do the job yet!) and cross-checked the contamination times with the available evidence. Our investigation was very detailed and thorough, and we are confident that we were able to uncover all the relevant information.

One of the most challenging aspects of the project was investigating the installations. It was a tough nut to crack because it required a lot of resources from various departments like engineering and maintenance and other firms that had built your clean room. However, we were determined to complete it quickly so, we rolled up our sleeves and got to work!

After conducting extensive research and analyzing vast data, we finally arrived at our first hypothesis. It’s important to note that our approach was not straightforward, and we didn’t identify a single cause of the problem. Instead, we discovered multiple factors that contributed to various contaminations. For example, there were gaps in the disinfection procedures and structural problems in one of the ducts responsible for circulating the air. The air leaked into the room unfiltered via a small crack with a poorly mounted HEPA filter, and we also identified several other minor deviations that, when combined, further contributed to the problem. Still, after weighing all the findings, we could formulate our hypothesis and CAPA plan.

Our team worked diligently to prepare a comprehensive CAPA plan that included investments, training, and monitoring procedures. We also conducted a thorough risk assessment of the findings to ensure that all aspects of the plan were carefully considered.

Following implementing a series of critical measures, particularly the main short-term actions identified, we proceeded to restart the previously shut-down room that had not been operational for over four months. As part of quality control procedures, we conducted a repeat media fill to ensure all the issues were fully addressed. This time, we were better equipped and more focused, having thoroughly analyzed and learned from our previous mistakes. Our newly established clean room qualification process and updated protocols ensured a successful media fill with flying colors.

Despite our challenges, we persevered and emerged victorious from a long and painful journey. Our dedication to the task enabled us to acquire expertise and knowledge that would have been impossible to gain through any other means or training. Our lessons helped other sites worldwide and ensured we were better prepared for next time. Moreover, the company benefited tremendously from this journey, as we learned a lot and created a group of global aseptic experts who can now help other sites faster, and, as always, you learn more from failure than from success.