A Better Approach To Aseptic Process Simulation For Lyophilized Products

Proposed 2020 revisions to EU Annex 1¹ with respect to aseptic process simulation (APS) for lyophilized products have prompted extensive discussions on best practices for process simulation of the lyophilization unit operation. This article serves to address these expectations and present a risk-based, holistic best practice approach for APS for lyophilized drug products.

Annex 1 revision excerpt (2nd draft, Sec. 9.35):

The process simulation test should imitate as closely as possible the routine aseptic manufacturing process and include all the critical manufacturing steps, specifically:

1. Process simulation tests should assess all aseptic operations performed subsequent to the sterilization and decontamination cycles of materials utilised in the process to the point where the container is sealed.
2. For non-filterable formulations, any additional aseptic steps should be assessed.
3. Where aseptic manufacturing is performed under an inert atmosphere, the inert gas should be substituted with air in the process simulation unless anaerobic simulation is intended.
4. Processes requiring the addition of sterile powders should use an acceptable surrogate material in containers identical to those used in the process under evaluation.
5. Separate simulations of individual unit operations (e.g., processes involving drying, blending, milling and subdivision of a sterile powder) should generally be avoided. Any use of individual simulations should be supported by a documented justification and ensure that the sum total of the individual simulations continues to fully cover the whole process.
6. The process simulation procedure for lyophilized products should represent the entire aseptic processing chain including filling, transport, loading, chamber dwell, unloading and sealing under specified, documented and justified conditions representing worst case operating parameters.
7. The lyophilization process simulation should duplicate all aspects of the process, except those that may affect the viability or recovery of contaminants. For instance, boiling-over or actual freezing of the solution should be avoided. Factors to consider in determining APS design include, where applicable

• The use of air to break vacuum instead of nitrogen.
• Replicating the maximum interval between sterilization of the lyophilizer and its use.
• Replicating the maximum period of time between sterilization and lyophilization.
• Quantitative aspects of worst-case situations, e.g., loading the largest number of trays, replicating the longest duration of loading where the chamber is open to the environment.

Both the 2020 revision and the current EU Annex 1 Manufacture of Sterile Medicinal Products (March 2009)¹ specify that "The process simulation test should imitate as closely as possible the routine aseptic manufacturing process." The words "as closely as possible" are of critical importance for lyophilization as certain aspects of the "routine aseptic manufacturing process" for lyophilization would adversely affect the media itself and/or microbial recovery. First, with freezing temperatures of -50°C or below and secondary drying temperatures as high as 60°C, the normal shelf temperature extremes during the lyophilization cycle are well outside the recommended storage temperature for the nutrient medias of 2 to 25°C.² Best practice is, therefore, to load the media filled containers on shelves precooled as close as possible to the normal product loading temperature, but within the recommended media storage range of 2 to 25°C.^{3, 4} For example, APS for a product normally loaded onto shelves precooled to -50°C would use shelves at 5°C (the target temperature for 2 to 8°C storage to avoid excursions below 2°C) throughout the simulation. Conversely, a product loaded onto ambient shelves (without shelf temperature control) would be subsequently controlled at a setpoint equal to the normal aseptic area room temperature, such as 20°C, throughout the rest of the simulation. Second, the typical 20 to 1,000 µbar (0.02 to 1.0 mbar) chamber pressures used during lyophilization would quickly boil away the media. To avoid boiling the media, the chamber pressure during APS must not drop below the equilibrium vapor pressure of the media (essentially water, 32 mbar at 25°C) at the loading temperature. Best practice, therefore, is to maintain the chamber pressure between approximately 100 and 200 mbar during simulation of sublimation and secondary drying.

The proposed 2020 Annex 1 revision further specifies that "The process simulation procedure for lyophilized products should represent the entire aseptic processing chain under specified, documented and justified conditions representing worst case operating parameters." A risk analysis for each step in the lyophilization process is presented below in Table 1.

Table 1: Relative risk of contamination by lyophilization process step.

a – While the risk to products stoppered under full vacuum is low, the risk to the APS remains high.

Recommended Approach And Rationale

The level of risk during loading is a function of the loading method. Fully manual loading, with human operators inside the Grade A loading area who come into direct contact with the trays or containers, must be considered high risk and the maximum load simulated in each APS. This may be achieved by loading a sufficient number of empty trays or trays with unfilled containers in addition to the media-filled containers to achieve the maximum commercial loading. Conversely, fully automatic loading, with no direct operator intervention (except allowed interventions), is considered low risk and would require only media-filled units and a sufficient number of empty, stoppered containers to fully load the last shelf with media-filled containers. Semi-automatic loading systems, where additional measures have been implemented to reduce the risk of fully manual loading, must be individually assessed and classified accordingly with a documented rationale for the APS design. As with the filling operation, corrective interventions permitted during commercial operations must be included in the challenge; specific interventions for each APS will be defined in the protocol but if product probes are used during commercial operation, this practice must be encompassed by the APS. Given that probed product containers are routinely rejected during normal production, media-filled containers with product probes must likewise be rejected after unloading and not incubated as part of the APS. Loading begins with the first opening of the chamber door and ends when the door is closed and/or the lyophilization cycle initiated.  During operations where the chamber door is closed (fully or partially to minimize airflow) periodically during filling/loading, the maximum number of door openings and the total time that the door would normally remain open with containers exposed to the room should be encompassed during the APS. The time the door is closed should be captured as part of the filling operation. Similarly, unloading begins when the chamber door is opened to begin unloading and ends when the last container has been sealed. Alternatively, if unloading and capping are performed as separate unit operations with an intermediate hold, the end of unloading may be defined as removal of the last container from the chamber, with the maximum allowable storage time captured as part of the capping operation.

Shelf temperature and chamber pressure limitations that are required to avoid damaging the recovery of contaminants (i.e., adversely impacting viability or recovery) during primary and secondary drying, or "chamber dwell," substantially alter the level of microbial risk as compared to the normal commercial process. In both cases, pre-sterilization of the lyophilizer assures sterility prior to loading. However, while vapor flow during normal product lyophilization provides a mechanism to transport organisms potentially present in the chamber into the containers, the only possible movement within the chamber during APS is slow evaporation of the media. Furthermore, while the upstream pressure control (gas injection) during the normal product process provides additional turbulence to transport organisms throughout the chamber during the cycle, the 100 to 200 mbar pressure required to prevent boiling of the media is well above the typical operating range of these systems such that nitrogen injection pressure control cannot be used during APS. With no mechanism to transport particulates into the containers during chamber dwell, the microbial risk drops to near zero and the length of the chamber dwell becomes immaterial. In addition, with some cycles lasting up to a week or longer, chamber dwells simulating the full cycle duration risk damaging media due to evaporation, particularly considering that the evaporation rate increases as chamber pressure decreases. Moreover, the higher chamber pressure required to avoid boiling the media results in a lower pressure differential between the lyophilizer chamber and the external environment such that the worst-case commercial conditions for microbial ingress during chamber dwell cannot be replicated in the APS.

Finally, the violent air turbulence during chamber aeration at the end of the cycle for both the normal commercial process and the APS may easily distribute any contamination present throughout the chamber and potentially into the media-filled units, particularly given that vapor is now flowing into the containers. Thus, aeration presents a high risk of microbial contamination during APS of the lyophilization process. Therefore, best practice to assure worst-case APS of the lyophilization unit operation is to pull the partial vacuum and then aerate to atmospheric pressure multiple times. As full vacuum would boil off the media, three partial vacuums are recommended.  While a short hold after

breaking vacuum and before the next partial vacuum adds value by allowing any disturbed particles to settle, holding the partial vacuum longer risks damaging the media by evaporation.  Note that while stoppering under full vacuum substantially reduces the risk to product containers during aeration, full aeration must be performed prior to stoppering during APS to avoid inhibiting microbial recovery due to lack of oxygen such that aeration remains the greatest risk of failure.

In summary, the greatest risk of microbial contamination during lyophilization is, by far, redistribution of any potential contamination by the turbulent airflow during aeration, particularly given that air is flowing into the containers. As full vacuum would boil off the media, three partial vacuums are recommended to assure a worst-case simulation. The steps in the recommended approach are listed in Table 2.

Table 2: Summary of steps in the recommended APS process.

Conclusion

This paper describes a best practice method for media fill simulation of the lyophilization unit operation. While previous literature presented a regulatory perspective for aseptic media fill qualification evaluating vial size and fill volume parameters,⁷ the method described in this paper uses a holistic  risk-based process analysis to ensure a clear worst-case APS based on sound scientific rationale and a comprehensive understanding of how the process limitations required to avoid damaging the media inherently alter the APS. This method also offers a clear operational advantage in that, by eliminating the chamber dwell during APS, the operational capacity of the lyophilizer (typically the capacity-limiting factor in the manufacture of lyophilized products) is improved.

References

1. Eudralex. "Annex 1 Manufacture of Sterile Medicinal Products." EU Guidelines to Good Manufacturing Practice Vol.4., 2008.
2. United States Pharmacopeia and National Formulary (USP 71-NF 36). Rockville, MD: United States Pharmacopeial Convention; 2016
3. FDA Guidance: Lyophilization of Parenteral (7/93). 2014.
4. Recommendation on the Validation of Aseptic Processes. Pharmaceutical Inspection Convention. January 2011
5. Bossert, KA. "Aseptic Process Simulations for Lyophilized Products: Points to Consider." SP Scientific Webinar. Mar 2017.
6. Nail, S and L Wegiel. "Best Practices in Pharmaceutical Freeze Drying." ISLFD Midwest 2015 meeting.
7. Vijayakumar V et. al. “A regulatory perspective: Rationale for hold duration of partially stoppered media filled vials.” Lyophilization World.

Contact The Authors:

• David A Hamilton: david.hamilton@merck.com
• Ted Tharp: ted.tharp@abbvie.com
• Orla McGarvey: orla.mcgarvey@lonza.com
• Martin Frei: martin.frei@lonza.com
• Michael Dekner: michael.dekner@takeda.com
• Shyam B. Mehta: shyam.mehta02@tevapharm.com
• Xiaodong Chen: xiaodong.chen@bms.com
• Nunzio Zinfollino: nunzio.zinfollino@merckgroup.com
• Stefan Schneid: stefan.schneid@bayer.com
• Josh Briggs: josh.briggs@biogen.com
• Melissa Schreyer: mschreye@its.jnj.com
• Deborah Hill: deborah.hill@biophorum.com