Virus Risk Mitigation in Cell Culture Media

The contamination of bioreactors with various infective agents such as bacteria, mycoplasma, and viruses is a potential risk to patient safety. Viruses have been the cause of multiple bioreactor contaminations in recent years (1). A number of biopharmaceutical companies have reported on production-scale bioreactor contaminations by small non-enveloped viruses such as minute virus of mice (MVM) or vesivirus (2). Many significant players within the industry participate in the Consortium of Adventitious Agent Testing (CAACB) to address this topic.

There are a range of methods available to manufactures that reduce the risk of cell culture contaminations by bacteria or mycoplasma; however, the contamination risk from viruses, specifically small non-enveloped viruses, is a much greater challenge to the biopharmaceutical industry even when using chemical defined media (1). Classical sterilizing-grade filters and even 0.1-micron membranes cannot prevent contaminations by small non-enveloped viruses.  

Virus contamination is more difficult to detect than contamination caused by bacteria (3). Failure to detect viruses can result in the contamination of the entire downstream process and even the final drug product. A number of companies have reported lost production lots due to contaminations with MVM or vesivirus (2). In all these cases, the root-cause analysis showed that the most likely source was the contamination of media components, such as salts, during storage at an early stage in the supply chain (4–6). Controls to prevent mice entering these warehouses have been inadequate. 

Due to the high stability of these non-enveloped viruses, they can survive for long periods of time. Even one infectious particle is enough to contaminate a whole bioreactor. This one particle would find the perfect replication conditions in the bioreactor environment. It is almost impossible to identify a single particle within 1000 L of media before it enters the bioreactor.
 

Impact of a contamination

The consequences of such a contamination event may be fatal for a patient (see Figure 1). If a company identifies a contamination, the plant can be shut down and subjected to extensive cleaning efforts. This could lead to drug shortages and patients not receiving lifesaving drugs (7). 

Figure 1: Escalation cascade for a virus contamination event. (Courtesy of authors)

Upstream virus clearance solutions

There is a demand for a new viral clearance concept for use in upstream processing. In the past, researchers have investigated technologies such as gamma irradiation, UV-C irradiation, or high temperature short time (HTST) for their ability to remove viruses from cell culture media. Bovine serum is gamma irradiated at external facilities that can handle large volumes (8). HTST is usually only economically feasible when preparing large volumes of media at high flow rates because of the large investment required to build customized systems (9). Unfortunately, not all media components are heat-stable and scalability of HTST system can be difficult. UV-C is limited by the flowrate that can be applied and hence has limited application when preparing large volumes of media. Furthermore, the data published on these methods show these technologies all have varying effectiveness in inactivating small non-enveloped viruses. Until now, filtration, which is a efficient technology that is especially effective for the small non-enveloped viruses, has not been widely adopted (10). 
 

Bottlenecks of downstream virus filters 

To achieve the desired robust virus removal, some companies have implemented virus filters used in downstream processing for their media preparation to minimize the risk of viral contaminations. Companies can effectively use these filters for virus filtration of media produced for perfusion bioreactors because the flow rates are low due to long production times. The capacities for media itself are quite high, with classical downstream processing (DSP) filters when filtering chemical defined media (up to 10.000 L/m² (11). If the possibility exists to filter these volumes over multiple days instead of a few hours, classical DSP filters are a perfect solution and economically feasible as low filtration areas are sufficient. Unfortunately, during classical batch filtration, this is not the case. Due to the risk of bacterial contamination, the media must be filtered within a minimum of 24 hours and ideally during a working shift. To increase the overall speed of filtration, larger membrane areas are necessary, but this approach is expensive and makes this scenario unrealistic economically.   
 

Economical feasible solutions

The industry needed to address the low flow rates and high cost of using virus filters designed for downstream processing application for cell culture media filtration for the virus filtration of media to be economically feasible. This was the basis for the first virus retentive filter membranes specifically developed for media filtration (10).
 

Methods

Retention performance

Each lab module with a filtration area of 5.0 cm² was challenged with a single organism in duplicate runs. In total, five organisms have been tested namely, MVM, Leptospira licerasiae, Mycoplasma orale, Acholeplasma laidlawii, and Brevundimonas diminuta.

As a worst-case model virus for small non-enveloped viruses, MVM, a single stranded, 18-26 nm, non-enveloped, ssDNA virus from the parvovirus family has been used. Following the current ASTM standard F838, a challenge level of > 107 organisms per cm² effective filtration area was applied for L. licerasiae, M. orale, A. laidlawii, and B. diminuta (12).  The composition of the growth medium for M. orale and A. laidlawii is according to Folmsbee et al. (13).

Membrane throughput testing 
The throughput of the membrane was determined using 15 different media from three different suppliers. All trails were performed at constant pressure of 2.0 bar. The total throughput was measured after four hours of filtration.

Results

Retention performanceTable I provides a summary of the retention data from testing performed with the new membrane. The membrane provides virus retention performance that is equivalent to those filters designed for use in downstream processing applications. Retention of the small non-enveloped virus MVM exceeded four logs. The analytical methods could not detect MVM in the filtrate. It is reasonable to expect that the membrane will retain agents larger than these small viruse that could potentially contaminate cell cultures. Data show that the new membrane not only successfully retains small viruses but can also provide a sterile filtrate when challenged with B. diminuta, different mycoplasma species, and L. licerasiae.  Companies have reported that this latter organism has contaminated cell culture by penetrating 0.1-micron media filters. 

Table I: Retentive capacities of the new virus retentive media filter, lab modules (5 cm2). CFU is colony forming units. LRV is log reduction value. EMJH is Ellinghausen-McCullough-Johnson-Harris media. HIMA is Health Industry Manufacturers Association.

Membrane throughput with different chemically defined cell culture media
The throughput of the membrane was determined using 15 different media from three different suppliers. Figure 2 shows that the volumes of media processed within four hours at 2-bar pressure varied widely depending on the media formulation. The Lonza Power CHO2, for example, blocked the membrane relatively quickly; however, the Lonza ProPER1 media did not appear to block the membrane at all. 

Figure 2: Throughput overview within four-hour filtration time with 15 cell culture media (CCM) at constant pressure 2.0 bar; virus media filter, lab modules. (Courtesy of authors)

For some of the commerical cell culture media, the use of an inline 0.1-micron filter increased the membrane throughput significantly (7). This was only successful with some of the media formulations and not with others. Protective agents, such as pluronics, reduce flux rates drastically (10). Reducing the pluronic concentration or filtering it before adding it to the media can increase filter throughput significantly. The industry should perform additional research to understand fully the influence that different media components can have on the blockage of media filters. It could use this research to design media with improved filterability characteristics without affecting its performance during cell culture. 

Nevertheless, the research performed shows that the new membrane developed typically filters approximately 1000 L/m² within four hours of commonly available cell culture media. This makes it an economically feasible method for the batch preparation of media while reducing the risk of viral contaminations of cell cultures.

Conclusion

In this article, the authors have reported on a new virus-retentive membrane that upstream process engineers can use to filter chemical defined cell culture media for risk mitigation. The authors have demonstrated that that this method is able to provide a greater than four-log reduction in small non-enveloped viruses. Furthermore, the membrane can retain bacteria and mycoplasma including Leptospira that could contaminate cell cultures. The high throughput of the membrane when used to filter a range of different chemical defined media enables biopharmaceutical companies to use the membrane in this upstream application without having a detrimental impact on the process economics. This membrane has the potential to be a crucial component of a risk-based approach to minimizing virus contamination events allowing the robust production of biopharmaceuticals.

References

1. P. W. Barone, PhD, “Lessons Learned from the Consortium on Adventitious Agent Contamination in Bio Manufacturing,” Viral Safety for Biologics, Cologne, June 21, 2016.
2. V. Bethencourt, Nature Biotechnology 27, 681 (2009), www.nature.com/nbt/journal/v27/n8/full/nbt0809-681a.html
3. O.-W. Merten, Cytotechnology 39, 91-116 (January 9, 2003). 
4. M. Moody, PhD, “MMV Contamination-A Case Study: Detection, Root Cause Determination, and Corrective Actions,” PDA/FDA Adventitious Viruses in Biologics: Detection and Mitigation Strategies Workshop, Bethesda, Maryland, Dec. 1-2, 2010.
5. Jim Skrine, “A Biotech Production Facility Contamination Case Study-Minute Mouse Virus,” PDA/FDA Adventitious Viruses in Biologics: Detection and Mitigation Strategies Workshop, Bethesda, Maryland, Dec. 1-2, 2010.
6.   Linda Hendricks, “Case Study of Apparent Virus Contamination in Biopharmaceutical Product at Janssen,” PDA/FDA Adventitious Viruses in Biologics: Detection and Mitigation Strategies Workshop, Bethesda, Maryland, Dec. 1-2, 2010.
7. B. Kleindienst, “Proof of Concept-The first virus retentive membrane for risk mitigation in upstream,” Viral Safety & Raw Materials for Biologics, Cologne, June 22nd, 2016.
8. G. Gauvin and R. Nims, PDA Journal, Vol. 64, No 5, 432-435, 2010. 
9. B. Hansmann, V. Thom, A. Manzke, “Contamination Risk Mitigation in Cell Culture Media Preparation,” Virus & TSE Safety Forum, Lisabon, June 9-11, 2015.
10. A. Meyer, “Risk Mitigation in Media Preparation-Current Possibilities and Future Trends,” PDA Virus & TSE Safety Forum, Berlin, June 5, 2013.
11. Thomas R. Kreil, “Virus Safety-A Look into the Entire Process Raw Materials, Upstream, Downstream,” European Upstream and Downstream Technology Forum Goettingen, September 8-10, 2014.
12. Health Industry Manufacturers Association (HIMA), Guidelines for Microbiological Evaluation of Filters for Sterilizing Liquids, 1982|ASTM F 838-05.
13. M. Folmsbee et al., PDA J Pharm Sci Technol., 68 (3) 281-96, May-June 2014.

Article Details

BioPharm International
Vol. 29, No. 10
Pages: 20–25

Citation

When referring to this article, please cite it as A. Manzke and B. Kleindienst, "Virus Risk Mitigation in Cell Culture Media," BioPharm International 29 (10) 2016.