“Sterilizing filtration is the process of removing microorganisms from a fluid stream without adversely affecting the product quality,” according to PDA Technical Report No. 26, revised 2008, Sterilizing Filtration of Liquids, page 2, paragraph 1. In that same report, on page 22, section 6.5, paragraphs 1 and 7, we are told, “Successful validation centers on the removal of a standard test organism (Brevundimonas diminuta ATCC® 19146™) at minimum challenge level of 107 cfu/cm2 of filter area.
Sterile filtration plays a critical role in patient safety and thus is highly regulated, being the last unit operation of a drug manufacturing process. Thus, it is a universal regulatory mandate that the performance of a critical sterile filtration step is validated in a process and product-specific manner. Table 1 provides a sample of regulations and industry guidance pertaining to validation of sterile filtration processes.
While clear guidance exists from PDA, FDA and EMEA on defining the components of a good sterile filtration process validation, one of the areas of less clarity relates to the length of time a filter can be used or the number of batches that can be produced with a single filter (Table 2). However it’s clear that while filter “re-use” is allowed, the responsibility resides with the filter user to execute a qualification and validation plan that includes worst-case consideration of process parameters under the extended use conditions.
The scope of “filter re-use” discussed here is sterilizing grade filters that are used for a campaign of multiple batches/lots of a single drug compound in these applications:
- Final bulk sterile APIs
- Terminally sterilized injectables
- Aseptically filled sterile injectables
- Aseptically filled sterile ophthalmics
There are a number of factors to take into account when considering validation of a sterilizing grade filter for use in multiple batches (of a single drug) or for filter use beyond one working day.
Economics of Re-Use
It is important to have clarity regarding the potential economic gain of re-using a critical consumable such as a sterilizing grade filter. Figure 1 demonstrates a simple calculation to determine the apparent potential savings relative to the product value. Clearly a process that produces a higher value product offers less economic gain for re-using filters.
Based on estimated inputs to this cost model, the perceived economic gain was estimated for several pharma manufacturing segments; it is evident that there is significant variability. (Note that these models show general assumptions based on our experience working with these processes.)
Potential Added Risk
Re-use of sterilizing filters has the potential to bring added risk. While risk mitigation can be achieved through effective validation, such validation will add complexity and cost to the overall process. Additional routine operations — cleaning, flushing, drying, storage, etc. — can also present additional risks to the filter, integrity, performance, sterility and ultimately the sterility of the product. Procedures must be carefully developed and executed to minimize these additional risks.
Integrity or sterility failure could lead to discarding or reprocessing of multiple batches of product when the filter is utilized for more than one batch. Potential causes for failure could include:
- Additional hydraulic stress over accumulated hours of use
- Additional thermal stress where multiple sterilizations are performed
- Chemical incompatibility, which could lead to failure over accumulated hours of use
- Membrane bacteria grow-through during long product wet hold periods between batches
- Additional physical stress from high pressure blow down operations to remove residual product and/or dry the filter membrane
Further, the added complexity of re-use brings forth a compliance risk; as re-use in critical sterile filtration processes is not common, it could bring added regulatory scrutiny.
You can expect that a regulator who sees re-use in a critical aseptic filtration process will be very interested in how the process was validated, qualified and operated. As such, before approaching the validation of re-use, a number of factors should be considered:
- Liquid filter re-utilization is quite rare in aseptic processing; although this practice is allowed by regulatory agencies, it is also discouraged, as is evident in the guidance documents.
- Any aseptic or sterile filtration application with re-utilization will likely be reviewed thoroughly by regulatory agencies for potential risks to the process, product, and/or patient.
- A well-designed validation strategy should be established to effectively mitigate the risks associated with filter re-use.
- The validation strategy for re-use applications should be developed on a case-by-case basis.
“Your response failed to ensure that products shipped to the U.S. market do not contain unacceptable residues from prior batches of different products that used the same filters. The inspection found inadequate validation rinse data to support that drug residue from the previous batch was removed. We are concerned of the possibility of cross contamination”
“According to the establishment inspection report you re-use sterilizing filters as long as they perform to the manufacturer’s filter integrity standards of [redacted] or for a maximum of your internal specification of 50 re-uses. We are concerned about the effectiveness of your filters after being re-used and autoclaved 50 times. We understand that you conduct a filter integrity test by the [redacted] method before and after each batch. However, in order to justify 50 re-uses, we would like to see bacterial retention validation studies using product both upon the initial use of the filter and after the 50th re-use.”
“Your response fails to address the major deficiency in the use of the (b)(4) system. Specifically, you failed to demonstrate that after cleaning, the equipment microbial bioburden and endotoxin levels met predetermined acceptable levels. Your firm did not support your cleaning and sanitization procedures with data supporting the system’s ability to effectively remove these contaminants. Your response provides reasons why the (b)(4) system did not need to be validated, rather than addressing the insufficiency of the data supporting the adequacy of your cleaning and sanitizing procedures.”
“Failure to validate the extended re-use period of (redacted) filters used for many different injectable product formulations and batches. Validation rinse data was inadequate to support that ingredient residue from the previous batch is removed.”
Based on our experience, we do not recommend re-use of sterilizing grade filters due to the significant risk of biological and/or chemical contamination to the process fluid. Examples of contaminations include, but are not limited to:
- Carryover of product from batch to batch
- Inadequate removal of cleaning agents
- Potential for microorganism to grow through the filter membrane
- Endotoxin from the breakdown of biological contamination
Below, we offer three case studies and demonstrate how the validation of unique re-use processes may be addressed to effectively mitigate risk and ensure regulatory compliance:
- Filter re-used for a campaign of multiple batches/lots with no cleaning and/or sterilization processes between batches/lots
- Filter re-used for multiple batches/lots with a cleaning cycle but no sterilization processes between batches/lots
- Filter re-used for multiple batches/lots with sterilization processes between batches/lots
Compatibility: It is recommended that for applications of re-use with prolonged total processing times that a full device compatibility test is conducted. Throughout a filtration process the hydraulic and chemical rigors of the process will be seen by a filter device, not a flat-stock membrane. The membrane within a filter device is often supported in a much different manner than a flat-stock disc in a holder and may also incorporate additional materials of construction that are critical to the filter device integrity and functionality. Thus, completing compatibility testing on the actual filter device will mitigate these risks by proving that the filter device can withstand the rigors of prolonged processing. It is important to look at multiple membrane characteristics during this test, because even a filter that passes a manufacturer’s integrity test specification could have its functionality compromised. In one example, it could be reduced flux due to membrane swelling. Evaluating the filter device with periodic measurements over time is also recommended, as this helps demonstrate suitability of the filter device for the prolonged application. A compatibility test of the filter device in the manner described will help support the decision to conduct a scaled-down bacterial retention test.
Bacterial Retention: If the filtration process does not include cleaning and/or sanitization of the filters between uses, the validation could be treated as a standard, scale-down (flat stock disc) long duration retention test where product conditioning and/or challenge are conducted for the same number of total cumulative hours that the filter is in use (first wet with product through final use). As with any well-designed bacterial retention test, product blow down or other methods of product recovery should be modeled in order to mitigate any potential risk of high differential pressure operation. Also, any extended hold periods should be included (test organism viability permitting) to ensure that the potential risk of bacterial grow-through will not result in non-sterile filter effluents when flow through the filter is resumed after an extended hold.
Extractables and Leachables: In this scenario, a standard extraction could be utilized where the filter’s pre-use conditions are modeled followed by extraction at worst-case processing temperature for, at a minimum, the full expected product contact time of cumulative use. The standard leachables risk assessment would apply.
Scenario #2: Filter re-use for multiple batches/lots with a rinsing or cleaning cycle but no sterilization processes between batches/lots. A process operated in this manner may or may not contain blow down steps to recover product and/or extended in-process hold times between batches.
Compatibility: Same as scenario #1 described above, except model any cleaning fluids (other than water) to the compatibility study.
Bacterial Retention: Same as scenario #1, except model any cleaning fluids (other than water) to the retention validation. However, if the volume of a water rinse is greater than 10% of the total product volume, it should be modeled in the study.
Extractables and Leachables: Same as scenario #1, except model or include rationale for any cleaning fluids (other than water) in the extractions. Cleaning solutions and process fluid would not be modeled by the same solvent for filter extraction; thus the process may require the completion of two or more filter extractions depending on the number of fluids used for filter cleaning. The potential risk of product or cleaning solution carry over must be addressed by the same considerations that would be taken for cleaning validation.
Scenario #3: Filter re-used for multiple batches/lots with sterilization process between batches/lots. A process operated in this manner may or may not include rinsing, cleaning, or blow down steps to recover product and/or extended in-process hold times between batches or drying. All special circumstances will be included in the process simulations for filter validation where they are deemed to result in a worst-case condition for the testing conducted.
Compatibility: For the same reasons offered in the discussion of scenario #1, it is also recommended that a full device compatibility test be conducted. Testing for this scenario, however, will include any processing between batches including cleaning or rinsing fluids, drying, and sterilization cycles.
Bacterial Retention: If there is a sterilization cycle in between each use, the cycle will need to be included in the retention validation. When simulating multiple sterilization cycles, a disc test is no longer the best test model. For example, the membrane may not be supported in a disc holder or small scale device in the same way that it would be in a process scale filter device, potentially resulting in differences in thermal and hydraulic stress. In addition, repeated exposure to differences in thermal expansion between membranes discs and stainless steel components in small scale systems make the membrane more apt to rupture in multiple cycles in this format, so multiple sterilizations should not be performed using the disc membrane.
Extractables and Leachables: The same rationale for extractables and leachables testing will be applied for this scenario. The worst-case would be to model only the cycles prior to first use. Pre-use flushes are not modeled for extractables studies.
In summary, properly validating filter re-use is complex but not impossible. It should be clear that one cannot expect to validate filter re-use with standard validation test services designed for single-batch production. Each re-use situation must be evaluated fully and a custom validation approach be designed to mitigate the potential process, product and patient risks. The filter user must weigh the perceived cost savings of filter re-use against the additional costs required to properly qualify and validate the process.
Ross Acucena is a regulatory consultant, Services & Solutions, at EMD Millipore. He can be reached at firstname.lastname@example.org. Randy Wilkins is Technical Consulting North American Team Manager, EMD Millipore. He can be reached at email@example.com.