Abstract
Mycoplasma are bacteria that can penetrate 0.2 and 0.22 μm rated sterilizing-grade filters and even some 0.1 μm rated filters. Primary applications for mycoplasma filtration include large scale mammalian and bacterial cell culture media and serum filtration. The Parenteral Drug Association recognized the absence of standard industry test parameters for testing and classifying 0.1 μm rated filters for mycoplasma clearance and formed a task force to formulate consensus test parameters. The task force established some test parameters by common agreement, based upon general industry practices, without the need for additional testing. However, the culture medium and incubation conditions, for generating test mycoplasma cells, varied from filter company to filter company and was recognized as a serious gap by the task force. Standardization of the culture medium and incubation conditions required collaborative testing in both commercial filter company laboratories and in an Independent laboratory (Table I). The use of consensus test parameters will facilitate the ultimate cross-industry goal of standardization of 0.1 μm filter claims for mycoplasma clearance. However, it is still important to recognize filter performance will depend on the actual conditions of use. Therefore end users should consider, using a risk-based approach, whether process-specific evaluation of filter performance may be warranted for their application.
LAY ABSTRACT: Mycoplasma are small bacteria that have the ability to penetrate sterilizing-grade filters. Filtration of large-scale mammalian and bacterial cell culture media is an example of an industry process where effective filtration of mycoplasma is required. The Parenteral Drug Association recognized the absence of industry standard test parameters for evaluating mycoplasma clearance filters by filter manufacturers and formed a task force to formulate such a consensus among manufacturers. The use of standardized test parameters by filter manufacturers, including the preparation of the culture broth, will facilitate the end user's evaluation of the mycoplasma clearance claims provided by filter vendors. However, it is still important to recognize filter performance will depend on the actual conditions of use; therefore end users should consider, using a risk-based approach, whether process-specific evaluation of filter performance may be warranted for their application.
Task Force Objective
To enable filter users to fairly evaluate individual 0.1 μm filter types for mycoplasma clearance based upon standardized test parameters, the Parenteral Drug Association formed a task force. The objective was to generate consensus test parameters, to be used by filter manufacturers, when rating 0.1 μm rated filters for mycoplasma removal. These consensus test parameters will be applied to filters of varying size and configuration, the test details of which will be defined by the manufacturer for each filter. For clarity, the objective was to agree on a set of test parameters, which are not necessarily worst-case parameters or perfectly representative parameters that would represent all conditions of use. We did not apply the test to any 0.1 μm rated filters because it was not our intent to evaluate the performance of any given or presumptive “0.1 μm rated” filter. Our intent was to set the parameters by which they would be tested and the test validated. Because filter performance will depend on the conditions of use, end users should consider, using a risk-based approach, whether process-specific evaluation of filter performance may be warranted for their application.
Introduction
“Mycoplasma” is often used to describe microorganisms from the class Mollicutes; however, within the taxonomic class Mollicutes, Mycoplasma also describes a specific taxonomic genera. Mollicutes (mycoplasma) are distinguished by their small size and lack of a cell wall, and as such they represent a unique challenge to removal by filtration. Initially, they were confused with viruses due to the ability to penetrate bacterial retentive filters (1) and later they were confused with “L-form” bacteria due to their lack of a cell wall (1⇓⇓–4).
With the advent of genomic sequencing, Mollicutes eventually received autonomous taxonomic status. Taxonomically, they are divided into ten genera distinguished by nutritional requirements, oxygen usage, metabolic capabilities, and their host preference (1, 5). Acholeplasma and Mycoplasma are two major parasitic genera of this class, each with strict host and tissue specificities, although some, such as Acholeplasma laidlawii, can live independent of a host (1, 6).
Mycoplasma are often difficult to culture due to specific and complex growth factor requirements, which are mostly unknown. To meet the complex nutritional demands, culture media for mycoplasma are generally undefined, containing mammalian tissue infusions, enzymatic digests of protein, yeast extract, and serum. As a general rule, Acholeplasma are more easily cultured than Mycoplasma, and have fewer nutritional requirements than other mycoplasma. Acholeplasma do not require cholesterol or saturated long chain fatty acids although they do require unsaturated long chain fatty acids (7).
A. laidlawii has been extensively used to study membrane lipids due to the lack of a cell wall and the relative ease of culture (8). As is true of other mycoplasma, A. laidlawii is dependent on its growth medium for supplying many of the compounds necessary to make new cells due to its limited biosynthetic capability. As a result, the nutritional composition (in particular, the fatty acid composition) of the medium strongly influences the composition of the cell membrane (7, 9⇓⇓⇓⇓⇓⇓⇓–17). Due to the influence of the culture conditions on the cell membrane, the culture media used to generate test cells is a relevant consideration with regard to the bacterial retention testing of filters (8, 18, 19).
Not surprisingly, bacteria will respond to culture conditions in a manner that will enhance their survival; environmental or nutritional stress can result in changes in cell size, cell density, cell morphology, cell growth rates, or other metabolic activity (20⇓⇓–23). For example, bacteria that survive starvation are typically reduced in size, grow more slowly (or not at all), and adapt better to adversity (22, 24⇓⇓⇓–28).
Possibly due to their small size and lack of rigidity (due to the lack of a cell wall), mycoplasma are demonstrably capable of penetrating 0.2 and 0.22 μm rated sterilizing-grade filters and even some 0.1 μm rated filters (18, 29). Primary applications for mycoplasma risk reduction filtration include large-scale mammalian cell culture media used in production of recombinant proteins or monoclonal antibodies, and bacterial cell culture media used for aseptic process validation by sterile process simulation or “media fill”. Common culture media components such as serum and enzymatic protein digests have been shown to be contaminated with mycoplasma, and mycoplasma are reported as common contaminants of cell cultures and cell culture media (30⇓–32). Increasing concerns with mycoplasma contamination of media raw materials led to the initiation of a workshop specifically addressing the issue of mycoplasma contamination by plant-derived peptones by the Parenteral Drug Association (PDA) in 2007 (32), and the development of a PDA technical report for addressing mycoplasma risk reduction filtration is currently in progress. The work described here was performed in support of that PDA Task Force.
Removal of mycoplasma by filtration requires the use of 0.1 μm or possibly tighter rated filters to provide greater removal efficiency than that afforded by 0.2 or 0.22 μm rated filters. Standard test parameters have not been defined for the rating of 0.1 μm filters. Consequently, each filter manufacturer has defined a 0.1 μm rated filter based on different test parameters and with test cells generated under different culture conditions. The key factor to understand is that microbial retentive filters (sterilizing-grade, mycoplasma-retentive, or otherwise) are not rated by filter manufacturers based on a direct measurement of pore size, but rather on the ability to retain an appropriate model organism (e.g., a standard bacterial strain) under given test conditions (which are usually based on ASTM F838-05, which applies to 0.2 μm filters) (33). In the case of 0.1 μm filters, until this report there has been no standard mycoplasma preparation for this purpose. For mycoplasma risk reduction, there is no regulatory requirement for end users to perform process specific evaluation of 0.1 μm rated filtration, but it can be a sound risk mitigation practice.
Filter manufacturers use bacterial retention test methods to characterize filter performance under controlled conditions during product development and as part of manufacturing lot release. It is important to recognize that the test conditions used to develop bacterial retention claims, while targeting extremes such as bacterial concentration and test pressures, cannot feasibly address all end user conditions. Consequently the filter manufacturer's bacterial challenge test is not intended to be representative of the end user's product or process conditions, and may not be as predictive of actual process removal efficacy.
Because a filter cannot be re-used after bacterial retention testing, filter integrity tests such as diffusive flow or bubble point serve as non-destructive methods for evaluating filter performance. During membrane development, a filter manufacturer will develop a correlation between bacterial retention and filter integrity to establish the performance boundaries of bacterial retention. The integrity test specification will be established based upon these criteria. Safety factors are typically built into this specification. During normal manufacturing, filters are integrity-tested prior to release by the filter manufacturer to confirm individual production filter performance. At use, the end user can perform an in situ integrity test and compare the results to the manufacturer's integrity specification to verify the filter is correctly installed and has not been damaged.
While filter manufacturers provide integrity test specifications and bacterial retention claims on their certificates of quality/analysis for sterilizing-grade filters, in the case of final sterile dosage form filtration processes it is the end user's responsibility to validate bacterial retention under actual product and process conditions to meet regulatory requirements for sterile drug products (or their relevant product or process) (33⇓⇓⇓–37). As yet, there is no equivalent requirement for process-specific evaluation of mycoplasma risk reduction filtration. However, filters are rated based on performance, and not based on an absolute measure of pore size. For diminutive-sized microorganisms, when present in the bioburden, end users may consider filtration process evaluation testing using an appropriate scale-down model, representative process fluids, and any test organism deemed relevant to their product or process. As a risk mitigation strategy, this, along with other risk reduction approaches, can help ensure the filter will perform as expected, under the conditions of actual use. Such studies may also be useful for investigations of the unlikely event of filtration process failures. In such filtration process evaluations, the filter material is tested for microbial retention using the worst-case formulation of the actual product (or an appropriate simulant) under worst-case conditions of process pressure and/or flow rate, process temperature, and any other relevant conditions. Worst-case conditions are interpreted to be those most likely to facilitate bacterial penetration.
Filter Rating and Efficiency
Mycoplasma clearance filters are generally described by filter vendors as having a 0.1 μm rating and may additionally be described as providing a titer reduction (TR) or a log reduction value (LRV). Prior to the consensus test parameters advanced in this paper, vendor-specific 0.1 μm ratings may have been established based on mycoplasma retention performance tests, on integrity test values, such as a bubble point or diffusive flow value, particle retention tests (such as latex beads), or some other manufacturer-defined parameter.
In mycoplasma challenge testing for manufacturer release or during filter development studies, mycoplasma-retentive filters are generally challenged with A. laidlawii, at a concentration of ≥1 × 107 colony forming units (CFU)/cm2 of filter surface area (comparable to ASTM-F838-05 recommendations, although this standard does not apply to 0.1 μm rated filters). Some mycoplasma filters may also have been tested with other mycoplasma species (e.g., Mycoplasma orale), as well as with the bacteria Brevundimonas diminuta.
Not all 0.1 μm rated filters claim 100% clearance of mycoplasma (no penetration at the specified challenge level and under the specified test conditions) and instead provide a TR or a LRV. TR is a measure of the degree to which a particular filter removes a microorganism under specified test conditions. TR is calculated as the ratio of the total number of bacteria used to challenge the filter divided by the total number of bacteria that passed through the filter:
The TR expressed as a base 10 logarithm becomes the LRV.
Mycoplasma Challenge Testing
In mycoplasma challenge testing, the test organism is inoculated directly into the challenge fluid and delivered to the test filter for a particular length of time, volume throughput, flow rate and/or pressure, and temperature. The test organism is suspended in the challenge fluid at a concentration that delivers a challenge level that meets or exceeds the 1.0 × 107 CFU/cm2 of test filter area. A sample of the influent challenge fluid is titered to determine the actual challenge level. The concentration of the test organism in the effluent is also determined, and the two are compared to evaluate the filter performance.
There are two possible outcomes to a bacterial challenge: (1) no penetration by the test organism occurs under the given test conditions, or (2) there is some degree of penetration under the given testing conditions. When penetration is detected, the TR or LRV is determined. When no penetration occurs (100% retentive under the given conditions), a TR or LRV may also be given with a greater than or equal to symbol (≥) to represent the maximum TR or LRV detected under the given test conditions.
Test Evaluation Criteria
Although it was not possible, and possibly not necessary, to develop a perfectly representative method, representing all possible end user manufacturing scenarios, it was important to ensure that the consensus test parameters would generate robust test conditions. Robust test conditions would ensure that a consistently penetrative test organism was used across multiple testing laboratories and under test conditions that would routinely result in the penetration of a 0.2 μm rated membrane filter. Demonstration of consistent penetration in this case refers to consistently demonstrating penetration (i.e., by at least one cell), not consistently demonstrating a set degree of penetration (i.e., TR). Penetration of a 0.2 μm rated filter confirms the small size, viability, monodispersion (dispersed and unclumped cells), and the overall penetrative ability of the test organism. Therefore a 0.2 μm rated filter is used as a positive control, a penetration control, similar to the use of a 0.45 μm rated penetration control filter in a sterilizing-grade filter challenge (as per ASTM-F838-05 recommendations) (33, 38, 39). As a result, the primary criterion for test evaluation was penetration of a 0.2 μm rated filter (the penetration control filter).
For clarity, the purpose of the penetration control filter is to demonstrate the ability of the test organism to penetrate a 0.2 μm rated filter. The penetration control for this test will be held to the same standard applied to the penetration control used when testing 0.2 μm rated filters. In a sterilizing-grade filtration test (0.2 μm rated), the penetration requirement is limited to the demonstration of penetration of a 0.45 μm rated penetration control filter. Penetration by a single cell is sufficient, and a minimum degree of penetration (a TR) is not required. In the case of rating a 0.1 μm filter (as described here), the requirement will be to demonstrate penetration of a 0.2 μm rated filter. As for the sterilizing-grade filter, penetration by a single cell is sufficient to demonstrate the penetrative ability of the test organism.
In an actual test, the penetration control filter must be tested in parallel with the test filter and serves to validate the test at the time it is performed. Lack of mycoplasma penetration of the 0.2 μm penetration control filter invalidates the test. The lack of mycoplasma penetration of the 0.2 μm penetration control filter may be due to (1) a lack of penetrative ability of the test culture, or (2) a high level of performance of the control filter. Unfortunately, either way, when it happens the test must be invalidated due to the inability to show the penetrative ability of the test culture.
It is important to understand that there is no upper limit to performance of any given 0.2 μm rated filter. Any individual 0.2 μm rated penetration control filter, of any lot, from any part number can, on occasion, perform well above the minimum requirements and demonstrate complete retention, even if the culture is actually highly penetrative. As a result, it is to the benefit of the test laboratory that the penetration control filter is pre-qualified prior to performing a challenge test. However, pre-qualification of the penetration control filter does not eliminate the risk of test invalidation due to a failure to penetrate a highly retentive filter; it only reduces the risk.
Establishment of Consensus Test Parameters
The task force established some test parameters for evaluation of presumptive 0.1 μm rated filters, by common agreement, based upon general industry practices without the need for additional testing (Table I). It was agreed that the culture medium and incubation conditions were critical elements to standardize. Standardization of the culture medium and incubation conditions involved collaborative testing in both (1) commercial filter company laboratories, and (2) two independent laboratories. Note that we did not apply the test to any 0.1 μm rated filters because it was not our intent to evaluate the performance of any given “0.1 μm rated” filter. Our intent was to set the parameters by which they would be tested (and rated) and how the test would be validated.
The Development of Consensus Test Parameters for Rating a 0.1 μm Rated Filter
The following consensus test parameters did not require testing to reach a consensus among the filter manufacturers:
The test organism: A. laidlawii ATCC® 23206™ (NC 10116; Strain PG8)
Challenge level (≥1 × 107 CFU/cm2 of test filter area)
The test pressure: 30 psid (207 kPa)
The challenge volume for a 47 mm disc: 200 mL
The challenge fluid: phosphate buffer (see Phase 2 Methods and Materials)
The positive control (penetration control): penetration of a 0.2 μm rated filter
The following test parameter did require testing to reach a consensus among the filter manufacturers:
The culture medium with which to generate test cells
The primary criteria for evaluation of the culture media was the ability to generate sufficient test cells and the ability to routinely penetrate a 0.2 μm rated penetration control filter. The following work was performed to evaluate several test culture media candidates with regard to A. laidlawii cell titer and penetrative ability. There were five candidate media formulations, four of which from four different filter companies and one from an independent laboratory.
In Phase 1 testing (the initial culture media screen), the five media formulations were tested at an independent laboratory (Independent Laboratory 1) with the assistance of the PDA Task Force Committee. The committee provided the culture media, test equipment, and technical knowledge related to integrity testing and bacterial challenge testing. The independent lab provided the inoculum and expertise with regard to mycoplasma culture and titer.
In Phase 2 testing, all five media formulations were evaluated in a round robin study at each of the participating commercial filter company laboratories. The consensus challenge fluid was chosen at this time and consisted of a phosphate buffer. The challenge fluid was chosen by the five filter companies based on what was practical for routine testing for all companies. The committee provided the culture media and the test strain, and each commercial filter company performed the testing using its equipment and technical knowledge related to integrity testing and bacterial challenge testing.
In Phase 3 testing, one of the five media formulations from the Phase 2 study was selected and provided to a second independent laboratory (Independent Laboratory 2), where it was formulated and used to generate three batches of test cells for bacterial challenge testing with A laidlawii. The single medium was selected based upon the lower TR (indicating the greatest bacterial passage) provided by cells cultured in this broth. Again, the PDA Task Force Committee provided the test equipment and technical knowledge related to integrity testing and bacterial challenge testing to assist the independent laboratory. The independent laboratory provided the inoculum and expertise with regard to mycoplasma culture and titer.
Phase 1
Phase 1 Methods and Materials
The objective of Phase 1 was to compare the ability of five media formulations to support growth of A. laidlawii while generating cells capable of penetrating a 0.2 μm rated filter. Five media formulations were tested by inoculation with A. laidlawii, titered after 24, 48, and 72 h incubation, and used to challenge a 0.2 μm rated 47 mm filter disc penetration control filter with a pressure vessel.
Materials
1.1. Organism: A. laidlawii ATCC 23206 (NC 10116; Strain PG8)
1.2. Inoculum: 10 mL volumes of a culture in tryptone soya broth (TSB, supplemented with ampillicin and phenol red) with a titer of ≥1 × 106 colony forming units (cfu/mL) stored at –70 °C
1.3. Culture Media
1.3.1. Medium 1: Formulation provided by Filter Company 1 (prepared by the Independent Laboratory 1)
1.3.2. Medium 2: Supplied complete by Filter Company 2
1.3.3. Medium 3: Supplied complete by Filter Company 3
1.3.4. Medium 4: Formulation provided by Filter Company 4 (prepared by the Independent laboratory 1)
1.3.5. Medium 5: modified TSB with phenol red and ampicillin, supplied complete by Independent Laboratory 1 (Oxoid cold-filterable & irradiated, CM1065) and prepared as per a proprietary formulation.
Culture Method
2.1. Broth Culture Initiation and Incubation: Prior to inoculation, each culture broth was incubated at 36 ± 1 °C for 24 h (± 10 min). Each culture broth was inoculated with 10 mL of freshly thawed stock culture and incubated at 36 ± 1 °C under static conditions.
2.2. The broths were titered at 24, 48, and 72 h incubation.
2.3. The broths were titered after serial 10-fold dilutions and plated as per the standard titer procedure of the independent lab. Culture plates were incubated at 36 ± 1 °C for 7 days before colony counting.
2.4. After incubation, a calculated volume of the culture broth (containing mycoplasma cells) was added to 0.1% Peptone broth (Oxoid, CM08660) to achieve a concentration sufficient to meet a minimum challenge level of 1 × 107 CFU/cm2 of effective filter area. The volume of the culture broth sample added to the peptone broth (challenge fluid) was calculated based on the titer of the individual culture broth to ensure that sufficient test cells were added to meet the minimum challenge level. If the titer of one culture broth was lower than another, then more of that broth was added.
Bacterial Challenge
3.1. A single 0.2 μm rated penetration control filter type was randomly chosen by the independent laboratory to use as the test filter for the bacterial challenge tests. All the Phase 1 tests were conducted with the same test filter type from a single lot to minimize variability due to the filter.
3.2. Each filter was challenged with 200 mL of challenge fluid at 30 psid (207 kPa).
3.3. The effluent was titered as per standard procedure at the independent laboratory.
3.4. All test filters were subject to pre- and post-challenge integrity tests.
Phase 1 Results and Discussion
Two criteria were used to evaluate the candidate culture broths: (1) ability to support growth of mycoplasma, and (2) the ability to generate penetrative cells. The results of the growth studies indicated that Medium 2 supported the highest growth relative to the other culture media, and Medium 1 supported the least amount of mycoplasma growth (of those that showed growth) (Figure 1). A 72 h time point was not available for Medium 3.
The results of Phase 1 trials. The growth of the mycoplasma A. laidlawii in five different culture broths after incubation at 36 ± 1 °C for 72 h (3 days).
After the growth study was performed, as summarized in Figure 1, two bacterial challenge trials with the five candidate broths were performed. The results of the bacterial challenge tests indicate that all the tested culture broths except for Medium 5 (which was not tested due to a failure to grow at the time of testing) generated cells capable of penetrating a 0.2 μm rated filter (Table II). In one case, the titer of cells generated with Medium 1 appeared to increase after passage through the filter; however, this was most likely due to the inherent variability of titer determination (±0.5 log) and not due to an actual increase in cell number.
Results of Phase 1 Trials. All the Tested Culture Media Generated Cells Capable of Penetrating a 0.2 μm Rated Penetration Control Filter
Summary of Phase 1: Four of five candidate broths supported growth of A. laidlawii and facilitated penetration of a 0.2 μm rated penetration control filter.
Phase 2
Phase 2 Methods and Materials
The objective of Phase 2 was to have five different labs compare the ability of the candidate media formulations to support growth of Acholeplasma laidlawii and generate cells capable of penetrating a 0.2 μm rated penetration control filter.
Materials
1.1. Organism: A. laidlawii ATCC 23206 (NC10116; Strain PG8).
1.2. Inoculum: vials containing 1 mL frozen cells, provided by commercial Filter Company 2.
1.3. Challenge fluid: Phosphate buffer was used as the challenge fluid (test fluid in which the test cells were suspended). It was composed of (pH adjusted to 7.1):
1.3.1. Sodium phosphate monobasic: 3.36 g/L
1.3.2. Sodium phosphate dibasic: 10.22 g/L
1.4. Culture Media
1.4.1. Medium 1: Formulation provided by Filter Company 1 (prepared by each testing laboratory)
1.4.2. Medium 2: Supplied complete by Filter Company 2
1.4.3. Medium 3: Supplied complete by Filter Company 3
1.4.4. Medium 4: Formulation provided by Filter Company 4 (prepared by each testing laboratory)
1.4.5. Medium 5: Unmodified TSB, purchased commercially and prepared as per the manufacturer's instructions (Oxoid cold filterable, & irradiated CM1065). This was the same TSB as used for the base of the modified TBS utilized in Phase 1, but without the addition of ampicillin or phenol red (and it did not require the use of a proprietary formulation).
Culture Method
2.1. Broth culture initiation and incubation: Each culture broth was inoculated and incubated as per the provider's/manufacturer's instructions. In some cases, the incubation period was extended to allow for more growth when insufficient cells were present.
2.2. The broths were titered as per each independent test site protocol.
2.3. After incubation, a sample of the culture broth (containing mycoplasma cells) was added to the phosphate buffer challenge fluid to achieve a concentration sufficient to meet a minimum challenge level of 1 × 107 CFU/cm2 of effective filter area.
2.4. The challenge fluid was also inspected for monodispersion to ensure a minimum of 80% monodispersion. Monodispersion was estimated by microscopic examination of the test cells using a Petroff Hauser counting chamber. The number of individual cells were counted and compared to the number of clumped cells, which were counted as a single unit. The number of single cells was divided by the total number of counted units to determine the percent monodispersion.
Bacterial Challenge
3.1. The bacterial challenge was performed identically as in Phase 1 with a single lot of a 0.2 μm penetration control filter and 200 mL challenge fluid at 30 psid (207 kPa). All test filters were integrity-tested pre- and post-challenge, and the effluents were titered as per standard procedure for each test laboratory.
3.2. The acceptance criteria for the bacterial challenges tests was that they meet the minimum challenge of 1 × 107 CFU/cm2 and a minimum of 80% monodispersion.
3.3. The results from three to five different test sites (five commercial filter companies) were averaged and compared. Each value provided by a given test site was an average of three tests performed at the site. Where available, five data points were averaged (representing a total of 15 tests) and when not, a minimum of three data points (representing a total of nine tests) were averaged.
Phase 2 Results and Discussion
Three criteria were utilized to evaluate the candidate culture broths: (1) ability to support growth of mycoplasma, (2) ability to generate penetrative cells, and (3) medium flow rate through the test filter. The average of four replicate stock titers is shown in Figure 2. All the candidate media generated stock titers within a similar range. The lowest stock titer (Medium 2) was 4.2 × 107 CFU/mL, and the highest stock titer (Medium 1) was 9.4 × 107 CFU/mL.
The results of Phase 2 trials. The average stock titer of A. laidlawii in five different culture media (four replicates). The error bars indicate one standard deviation.
The average TR overall is shown in Figure 3. Each test laboratory performed the test in triplicate and provided an average value to the task force. These averaged values were then averaged together to generate Figure 3 (thus the bars represent the average of 12 to 15 tests). The lowest TR (most penetration) was provided by cells cultured in Medium 3, and the highest TR was provided by cell cultured in Medium 5. In two cases (for a total of 6 challenge tests), cells generated in Medium 5 failed to penetrate the test filter.
The results of Phase 2 trials. The average TR provided by test cells of A. laidlawii cultured in five different culture media. Each test laboratory performed the test in triplicate and provided an average value to the task force. These averaged values were then averaged together to generate the value shown above (thus the bars represent the average of 12 to 15 tests). The most penetrative cells generated the lowest TR. The error bars indicate one standard deviation.
The average flow rate obtained in the mycoplasma challenge tests is given in Figure 4. Each test laboratory performed the test in triplicate and provided an average value to the task force. These averaged values were then averaged together to generate Figure 4 (thus the bars represent the average of 12 to 15 tests). The flow rates obtained were similar and within expected values.
The results of Phase 2 trials. The average flow rate of mycoplasma challenge tests using a phosphate buffer as the challenge fluid and test cells generated in five different culture media. Each test laboratory performed the test in triplicate and provided an average value to the task force. These averaged values were then averaged together to generate the value shown above (thus the bars represent the average of 12 to 15 tests). The error bars indicate one standard deviation.
Summary of Phase 2: All candidate broths supported the generation of similar stock titers and similar flow rates (with similar variability). The average TRs resulting from the use of cells generated in Medias 1–4 varied less than that of Medium 5. Because Medium 5 generated the least penetrative cells, and in six cases generated cells that did not penetrate the 0.2 μm rated penetration control filter, it was no longer considered. Of the remaining four culture broths, Medium 3 was selected due to the lower TR provided by cells cultured in this broth. The composition of Medium 3 consists of the following (to be called Mycoplasma Task Force Broth):
1) 20 g/L mycoplasma broth base (beef heart infusion broth), composed of
Beef heart infusion (2.0 g/L)
Pancreatic digest of casein (7.0 g/L)
Beef extract (3.0 g/L)
Yeast extract (3 g/L)
Sodium chloride (5.0 g/L)
2) 25 g/L yeast extract
3) 100 mL/L of heat-inactivated horse serum (55 °C for a minimum of 30 min)
Phase 3
Phase 3 Methods and Materials
The objective of Phase 3 was to have a single independent lab document that the cells generated using the selected media would consistently penetrate several different 0.2 μm rated penetration control filters.
Materials
1.1. Organism: Acholeplasma laidlawii ATCC 23206 (NC10116; strain PG8)
1.2. Inoculum: Generated on-site by Independent Laboratory 2
1.2.1. Challenge fluid: as described previously in Phase 2 testing
1.3. Culture Medium and Incubation
1.3.1. Mycoplasma Task Force Broth formulation provided by the PDA Task Force and prepared in three batches by Independent Laboratory 2
Culture Method
2.1. Broth culture initiation and incubation: Each batch of culture broth was inoculated and incubated at 37 °C for three days.
2.2. The broths were titered as per Independent Laboratory 2 site protocol.
2.3. After incubation, a sample of the culture broth (containing mycoplasma cells) was added to the phosphate challenge fluid to achieve a concentration sufficient to meet a minimum challenge level of 1 × 107 CFU/cm2 of effective filter area.
2.4. The challenge fluid was also checked for monodispersion to ensure a minimum of 80% monodispersion.
Bacterial Challenge (a Schematic Diagram of the Test Equipment Is Given in Figure 5)
3.1. Three different, commercially available, 0.2 μm rated penetration control filter types were randomly chosen by the PDA Task Force Committee to use as the test filters for the bacterial challenge tests to demonstrate that the selected media produced cultures capable of penetrating different 0.2 μm filters.
3.2. Each filter was challenged with 200 mL of challenge fluid at 30 psid (207 kPa).
3.3. The effluent was titered as per standard procedure of Independent Laboratory 2.
3.4. All test filters were subject to pre- and post-challenge integrity tests.
Schematic diagram of the mycoplasma challenge for Phase 3 testing.
Phase 3 Results and Discussion
The objective of the final phase of the testing was to confirm penetration of the 0.2 μm penetration control filter using: (1) the chosen test parameters, and (2) the chosen culture medium #3 for the generation of test cells. Three batches of test cells were generated and three different 0.2 μm rated filters were tested with each batch of test cells. This was to confirm that the selected test parameters would provide a robust challenge to a 0.1 μm rated filter by demonstrating passage through a 0.2 μm rated penetration control filter. All three batches of test cells penetrated all three 0.2 μm rated test filters (Table III). All test filters passed pre- and post-challenge integrity tests.
The Results of Phase 3 Trials. The Results of Bacterial Challenge Tests of Three Different 0.2 μm Rated Penetration Control Filters with Three Batches of A. laidlawii Test Cells Generated Using the Selected Test Media
Conclusions
The goal of the Parenteral Drug Association–endorsed Mycoplasma Task Force was to identify consensus test parameters for the filter manufacturers' rating of 0.1 μm filters to be used for mycoplasma clearance. The consensus test parameters include some criteria based upon common industry practices, eliminating the need for additional testing. The culture medium and incubation conditions (described below) were standardized.
The consensus test parameters are:
1) The test organism: Acholeplasma laidlawii (NC10116; Strain PG8)
2) Challenge level (1 × 107 CFU/cm2)
3) The test pressure: 30 psid (207 kPa)
4) The challenge volume for a 47 mm disc: 200 mL
5) The challenge fluid: phosphate buffer
6) The positive control (penetration control): penetration of a 0.2 μm rated filter
7) Mycoplasma Task Force Broth for generating test cells
Ultimately, the ability to consistently penetrate a 0.2 μm rated control filter was the primary criteria for evaluation of the culture medium and incubation conditions. All candidate broths supported the generation of similar stock titers and similar flow rates. Medium 3 was selected due to the lower TR through a 0.2 μm rated filter (the penetration control for the test) provided by cells cultured in this broth, indicating a robust ability of the test cells to penetrate the control filter under our defined test conditions. These mycoplasma preparations will ensure a robust test of a 0.1 μm rated filter by a filter manufacturer, using the test parameters outlined here:
The temperature of incubation: 37 °C
The duration of incubation: 3 days
The composition of the Media 3 consists of the following:
20 g/L Mycoplasma broth base (Beef heart infusion broth)
25 g/L yeast extract
100 mL/L of heat inactivated horse serum (55 °C for a minimum of 30 min)
It is recognized that the culture media described here is not a defined culture media. Mycoplasma broth base (or beef heart infusion broth), yeast extract, and horse serum are not defined components, so there will be some variability in the media composition—not just between laboratories but also from batch to batch within a laboratory—due to the inherent variability of these undefined components. Alternatively, cultivation in TSB does not require these undefined components and would ensure a more consistent composition of the culture media; however, cultivation in TSB also led to a noticeable absence of penetrative cells, which was not useful to ensure a robust filter challenge test.
Ultimately, the requirement that the culture media generate penetrative cells was of more importance than the desire for a consistent media composition. So, although the culture media is not defined and will vary in some aspects, cultivation in this media consistently ensured the production of highly penetrative cells. As with the 0.2 μm penetration control filter, each testing laboratory will need to pre-qualify each component of the media with regard to bacterial growth. It is to the benefit of the test laboratory that the media components are pre-qualified prior to performing a challenge test. However, pre-qualification of the media does not completely eliminate the risk of test invalidation due to a failure to penetrate even a minimally retentive filter; it only reduces the risk. As previously stated, in an actual test, the 0.2 μm penetration control filter must be tested in parallel with the 0.1 μm test filter and ultimately serves to validate the test at the time it is performed.
- © PDA, Inc. 2014
