Microbiological Control for Affinity Capture Chromatography Processing: An Industry Perspective

Purpose of Industry Survey

The purpose of this survey was to develop best practices for microbiological control of the affinity capture chromatography unit operation:

• To outline historical microbiological concerns for this unit operation (i.e., source of microbial events, type of isolates recovered, etc.).

• Determine likely sources of microbial contamination.

• To focus on the material, environmental, equipment, and storage controls required to mitigate microbial ingress and biofilm formation in chromatography equipment.

• To list the potential remediation measures that may be taken during a microbial event.

BPOG Survey Method and Participation

All companies surveyed are biopharmaceutical manufacturers of both clinical and commercial cell culture products and are located predominately throughout the U.S. and Europe. Up to 14 respondents from drug substance manufacturers participated. Participants included subject matter experts from manufacturing, microbiology, and quality assurance. The survey was performed in a blinded manner, coordinated by a representative from the BPOG organization. The summary results are reflected in this document. The survey was initiated with general questions relating to chromatography and application within the manufacturing platform. The survey focused on historical events relating to microbiological control, with emphasis on potential points of microbial ingress, remediation procedures, and prevention controls. The recommendations for best practice from industry are included, comprising material, equipment, and operational controls.

Survey Results & Discussion

The application of affinity capture chromatography within the industry group is summarized in Table I.

Manufacturing experience has determined that bioburden control is a key component of affinity capture processing. Table II illustrates the primary concerns.

The detail of specific bioburden control concern was further elucidated by the participants in this survey. As outlined in Table III, water-borne species and typical environmental isolates are the predominant isolates detected. Detection of the contaminants was readily evident in the routine process monitoring applied to equilibration effluent and product pools, and less frequently from swab surface monitoring of chromatography equipment. The isolate types detected were closely correlated to the confirmed or putative point of ingress. Of particular interest was the recurring detection of Stenotrophomonas maltophilia. These isolates were detected across multiple facilities and companies, with no commonality other than resin type and, in some cases, the sanitization strategy. Equipment hardware design and soft part failure were common causes of concern; these may be mitigated through use of pre-packed columns. Weak and ineffective storage solutions lead to contamination when applied in situ or when in storage out-of-column (e.g., for longer periods between campaigns). It is also acknowledged that resin material is not free of microbial species, with typical incoming raw material specifications at <100 CFU/mL.

The response to each event was dependent on the scope and extent of contamination. The remediation efforts included extensive sanitizations, and also included column re-packing, equipment cleaning, and so forth. In some situations the resin material had to be discarded.

The experience of the survey participants enabled the collation of guidelines or best practices to ensure effective control of the resin material such that microbial ingress could be effectively prevented. These controls are summarized in the following section.

Best Practices for Affinity Capture Resin Storage

Resin is stored both in and out of column. While the majority of respondents unpack and store offline when not in use, three of the (14) respondents maintained a preference to store columns in situ between campaigns. Two respondents do not apply resin storage at all and discard resin at the end of the manufacturing campaign.

Where storage conditions apply, the majority of unpacked resin is stored in carboys at 2–8 °C. The same storage buffer is typically applied for in situ and out-of-column storage. Storage solutions such as 18–20% ethanol (buffered solutions may apply) and 1–2% benzylalcohol at 2–8 °C (pH 3.2 or 5.0) are applied, based on compatibility with the resin, vendor guidance, and facility requirements. It was noted that the concentration of the solution in storage can be dependent on the unpacking procedures, the containment provided during storage, and the storage duration. Sporeformers are well known to persist in 20% ethanol solutions, with some noted observations of microbial persistence in pH 5 benzyl alcohol. Of the 10 respondents, four provide microbiostasis data to support resin storage.

Where resin storage applies, validation of the effectiveness of the storage conditions is a regulatory expectation. Concurrent resin lifetime studies include an evaluation of resin storage, which may apply both in situ between batches, and out-of-column storage for longer term storage. Microbiological data is typically requested to support maximum resin storage durations.

In almost all cases (10 of 12 respondents), the effectiveness of resin storage is verified each time, through in-process control testing of the resin storage solutions or slurries prior to further processing. These in process controls may include one of the following:

• – through a blank elution prior to further processing,

• – directly on the resin slurry/test of slurried resin before column packing,

• – routine bioburden testing after sanitization of each column packing,

• – equilibration of the first cycle,

• – sampling of post-storage flush solution.

The benefit of these in-process controls was not evident in all cases, however, as only two respondents detected an issue with the resin prior to use. Due to the nature of downstream operations, it is not always operationally feasible to wait on bioburden data from the packing operation before further processing. The application of rapid microbiological methods would provide substantial benefit in this manufacturing step.

Short-term storage also applies within the chromatography column between batches and between cycles of individual batches. Storage controls and solutions used for such short-term intervals varied greatly among the respondents. For example, resin is sanitized and stored in 20% ethanol between individual cycles. This strategy incorporates a parallel sanitization and storage of the gradient cart equipment (i.e., chromatography skid) used for processing. Another strategy involves a defined sanitization and storage frequency, based on the number of cycles/batch or the processing duration (sanitization and storage after every 24 h). In these circumstances, the resin is regenerated between cycles and maintained in equilibrated solution until a defined time has elapsed or until each cycle of the particular batch has been processed across the unit operation. Continuous processing ensures no storage between chromatography cycles. A recommended bioburden control measure is to ensure that neither the packed resin nor the gradient cart equipment is stored in equilibrated solution for longer than 24 h. Where required for chromatography, gradient carts are stored in ethanol or caustic solutions (≥0.5N NaOH where possible). These microbiological controls are verified as a component of commercial-scale process validation of the unit operation.

Best Practices for Control of Process Inputs

A key control to ensure that the packed resin is effectively protected for microbial ingress is control of process inputs. Contaminated charge/load solutions was highlighted as a common concern in biopharma. The nutrient-rich, cell-free media may provide low-level bioburden, which can colonize the resin bed, making it potentially difficult to treat with routine sanitization solutions, given the niche spaces provided in a compressed resin bed. On-skid bioburden control filtration is applied as the primary measure of control for this reason. While guard filters may be used to protect the resin within the column, it is generally recommended that the load solution is subject to sterilizing-grade filtration prior to loading to the Protein A column. Filtration of incoming process and cleaning buffers may also be enabled in certain situations. Once process buffer connections have been established, avoid scenarios that require reattachment, as this increases the need for further manual interventions. These controls are accompanied by use of hygienic connections, aseptic technique, and effective training.

Chromatography Equipment and Soft Parts Management

Process equipment and system components should be designed to minimize the potential for microbial contamination. Because common water-borne isolates have a proclivity to form biofilm in difficult-to-clean areas, the equipment should be designed and qualified to ensure effective cleaning and removal of residue and excess water rinsate. Ensure line slopes are provided for maximum drainability. Difficult-to-clean locations (e.g., inlet ports and valves) should be thoroughly inspected to ensure effective cleaning. The equipment should be designed to minimize any open operations.

Gradient carts utilize a variety of fixed and disposable components and are generally not subject to steam-in-place prior to use. Particular emphasis should be provided to the design of efficacious cleaning regimens to mitigate the risk of biofilm formation on the equipment surfaces. The general guidelines for equipment cleaning and soft parts management are thoroughly described in PDA Technical Report No. 69—Bioburden and Biofilm Management in Pharmaceutical Manufacturing Operations (1). Contamination events have been described in which cleaning steps were not performed with sufficient flow rate to remove air, leading to ineffective application of the sanitizer. Effective cleaning is ensured when equipment flow path and flow rate applied for sanitization is optimised and validated for microbial control. Following cleaning, gradient cart design principles should allow for complete draining, which prevents standing water. This is especially important in situations where gradient carts are stored dry. Where gradient carts are stored in bacteriostatic solutions following cleaning, removal of excess cleaning rinsate is also important.

Cleaning validation of ancillary equipment is also important. Effective validation for cleaning of empty columns, slurry tanks, resin storage containers, and packing skids has been demonstrated to improve microbial control for affinity capture chromatography operations.

Robust preventative maintenance procedures are necessary for this equipment. Regular inspection of soft parts for damage are required, with a change-out frequency which is based on thorough risk assessment of process use, contact with sanitization solutions, steam, and so on. Because misalignment of gaskets and failed control valve diaphragms can lead to ineffective cleaning and biofilm formation, inspection of seals and gaskets for leaks is required, with pressure testing of column seals recommended prior to use. Use of double-sealed O-rings provides additional protection to the process flow paths. Specialized components such as chromatography column gaskets/frits used for resin support may be subject to re-use and offline storage. Specialized storage and return-to-service procedures should be developed for such components.

Effective Resin Sanitization

Following a satisfactory resin pack operation and closure of the column, resin sanitization is utilized as the primary measure to control bioburden. The sanitization solutions applied vary across the companies surveyed and are largely resin-dependent or recommended by resin vendors based on compatibility studies. Typical examples include low concentrations of NaOH in NaCl-buffered solution (typically for MabSelect); however, guanidine HCl followed by acetic acid/benzyl alcohol or acetic acid/ethanol are also used.

Sanitization is applied in almost all cases after column packing and prior to the first cycle of each batch to be processed; however, thereafter the frequency of sanitization varies based on resin type and application in manufacturing. At least five of nine respondents sanitize the affinity capture resin after each cycle has been processed. The remaining respondents apply a sanitization schedule in line with the start and end of the processed batch (i.e., cycle number–dependent) or in a time-dependent manner (e.g., at least once per day during processing). The method of application varies, with almost half of the respondents applying the sanitization solution in up- and down-flow mode. This is recommended particularly following an intervention (e.g., column packing). The contact time with the sanitization solution varies greatly (15–120 min) and is largely applied without a static hold.

Each sanitization solution will have its advantages and limitations. For example, it is well known that low concentrations of NaOH are only effective for vegetative species, as Bacillus spp. have been observed to persist without additional intervention. Based on historical observations, many of the respondents have developed remediation measures for a suspected bioburden event.

The sanitization regimen is typically developed and assessed initially through laboratory-scale evaluations. However, the effectiveness of sanitization for microbial control can only be verified through continuous process verification at commercial scale. Based on these considerations, general best practice guidelines for resin sanitization include the following:

• The frequency of sanitization is important for microbial control. Column steps that do not have a frequent cleaning tend to have many more microbial events.

• The application of an effective regeneration (i.e., strip) step may enhance the effectiveness of sanitization step.

• For larger columns, ensure sufficient flow rate to ensure radial diffusion (via the column flow plate) throughout the column.

• For larger columns, ensuring homogenization of the resin during the packing helps to prevent channeling or other sources of uneven flow.

• Applying sanitization in up- and down-flow mode is known to enhance dispersion of sanitizer within compressed resin bed, while also helping to remove entrained air.

• A static hold allows for static diffusion of the sanitant throughout the compressed resin bed.

• Develop remediation measures for a suspected bioburden event (see below).

Effective Remediation Measures for a Suspected Bioburden Event

Detection of bioburden in equilibration effluent or product pools during successive chromatography cycles is likely to indicate a microbiological trend that requires effective corrective measures to restore control. While each event and the detected species will vary in each situation, almost half of the respondents have proactively determined remediation measures to be deployed. Examples of such measures are outlined in Table IV. The compatibility of the resin with stronger solutions is an obvious prerequisite. To facilitate such measures, a number of respondents indicate that laboratory-scale lifetime studies incorporate such remediation measures. While measures should be taken to ensure a representative challenge, it is acknowledged that these laboratory-scale studies cannot predict every manufacturing event over the lifetime of the resin, and also do not have significant breaks between every three to five cycles as required in traditional manufacturing. Remediation at commercial scale therefore needs to determine the steps necessary based on systematic risk assessment to ensure consideration of both the performance and lifetime of the resin and the requirement to re-establish microbial control. Note that in some cases, bioburden events are not easily resolved using these methods. In some situations, resin was unpacked and discarded.

Best Practices for Affinity Capture Packing and Unpacking Operations

A majority (7 out of 11) of the respondents identified column packing using fresh or re-used resin as the processing operation that is more susceptible than others to microbial ingress. Column packing is an open operation, where the resin is slurried in a buffer solution before being packed into a clean column. Companies have developed procedures to minimize these risks, with particular emphasis on limiting the time required for packing and by applying a post-packing sanitization activity. Additional recommendation is to provide for bacteriostatic solution storage for the resin if there is a delay in packing or when packing performance analyses need to be repeated.

The resin unpack process is also an open process. Where resin is subject to storage, the decanting operation to storage containers may be required. The majority of respondents perform these activities using a manual or semi-automated process. A packing skid is usually required. The common procedures applied to enhance control of these activities are as follows:

• Perform column packing and unpacking activities under cleanroom controls. A minimum Grade D classified area should be provided for this open operation.

• Where possible, provide a dedicated area.

• Close up the unit operation as much as possible.

• Gowning controls should include full coverall and sterile gloves/gauntlets and face masks, to minimize risk of personnel contamination during open steps.

• Limit the number of personnel in the area during the operation. Ensure personnel have specialized training and are trained in good aseptic technique.

• Limit the time required to conduct activities.

• Where possible, pack the resin in process solutions that limit microbiological growth (e.g., use storage buffers). If this is not possible, limit the time the unpacked resin remains within solutions that permit microbial growth.

• Utilize autoclaved components where possible.

• Verify for bioburden in a rinse sample after empty column cleaning.

Conflict of Interest Declaration

The authors declare that they have no competing interests.