Assuring sterility with ASTM F 838-83

Theodore H. Meltzer

Theodore H. Meltzer is principle of Capitola Consultancy.

,
Maik W. Jornitz

Pharmaceutical Technology Europe

Pharmaceutical Technology Europe, Pharmaceutical Technology Europe-12-01-2008, Volume 20, Issue 12

Although realized in 1983, ASTM's standrad test method for determining bacterial retention of membrane filters used for liquid filtration is still hugely beneficial to today's pharmaceutical industry.

Where possible, sterility of medicinal products is assured by heat sterilization. However, this cannot be used with every pharmaceutical product as it can alter or degrade the drug — particularly biologicals. In other instances, the gram-negative microbial bioburden level within the product is too high, meaning that heat sterilization would cause an undesirable concentration of endotoxin.

Microporous membrane filters were invented in 1918 and came to use in the early 1930s.1 These filters either created a sterile effluent or a bioburden level reduction to a point of no concern in regard to any potential endotoxin contamination. Membrane filters in those days could not be standardized or reliably compared in consideration of retention performances. Instead, filter manufacturers performed microbial tests under different challenge conditions, and with organisms grown under diverse conditions and with different culture media. The lack of a standard made it impossible for the filter user and regulatory authorities to determine which filter should be designated the sterilizing grade filter.

True comparability and performance analysis could only be performed when a standard microbial challenge test was defined that detailed:

  • Organism type (ATCC) to be used

  • Growth conditions and media composition

  • Challenge conditions

  • Test and sampling set-up.

The first steps towards establishing a standard were made with a document from the Health Industry Manufacturers Association (HIMA),2 which was later developed into ASTM standard F 838-83 and is now ASTM standard F 838-05.3,4 The focus on this standard for microbial testing led the industry to optimize sterile drug processing using filtration sterilization. The present biopharmaceutical interventions of fermentation-based drugs in medical treatments are the most recent consequences of ASTM standard F 838-05.

A quarter of a century later, the standard still serves as the bedrock of organism quantification by which the efficacy of drug sterilization by filtration is determined.

On the go...

History

The first sterilizing grade filters were rated 0.45 μ and were used to remove yeast, mould and bacteria from beverages and pharmaceutical fluids. The definition of pore size rating commonly used is not derived from direct measurements, but from the rates of flow of fluids, air or water, through the filters. In filtration practices, however, emphasis is so heavily on particulate retention that it is sometimes popularly, but erroneously, believed that filters are classified according to the actual physical dimensions of their pores.

Much ingenuity has been exercised in developing methods for measuring the size of filter pores. Flow measurements of both air and water have been widely used and resulted in an ASTM standard procedure (Figure 1).5–8 Other tests including mercury and/or air intrusion under high pressure have been employed and,9,10 more recently, vapour pressure analysis.11 New mathematical insights have also been brought to bear.12,13

Figure 1

A first step towards a challenge test definition was occasioned by an incident of bacterial filter penetration. In 1967, Bowman et al. published the discovery of a smaller, water-born organism,14 which penetrated the 0.45 μ-rated filters that were then considered the sterilizing grade membranes. This finding resulted in the development and universal use of 0.2 (0.22) μ-rated membrane filters that could retain the smaller organism. The organism was identified as Pseudomonas diminuta, but has now been reclassified as Brevundimonas diminuta.

As with 0.45 μ-rated filters, the 0.2 μ rating was not standardized and multiple different challenge tests were again published by filter manufacturers.

Sterilizing grade filters and their pore size ratings became critical tools within the industry and were, therefore, required to be tested accordingly. Any microbial breakthrough caused by an ill-defined rating could be detrimental to patient safety. Such occurrences were guarded against by assays enabled by the ASTM standard.

The industry took action against the lack of a standard challenge test for sterilizing grade filters by forming, in combination with the filter manufacturers, HIMA. It established a standard microbial challenge test for sterilizing grade filters rated at 0.2 (0.22) μ. This was the first time the challenge organism, challenge level, growth media and filtration conditions, as well as challenge conditions, were defined. Another beneficial outcome of the efforts of this group was the term 'log reduction value' (LRV) as a description for the microbial removal efficiency of a filter. A filter with an LRV of seven is capable of reducing the organisms in the feed stream by seven orders of magnitude.

From that point onwards, the HIMA test methodology was used by filter manufacturers as the challenge test for 0.2 (0.22) μ-rated sterilizing grade filters, and as a verification method attesting that the filter will retain the challenge organisms at a level of 107 /cm2 of filtration area. Nevertheless, it only became an official standard when ASTM used its committees and membership to refine and convert the HIMA standard into ASTM standard F 838-83 in 1983, creating a solid, scientific benchmark. This standard was reviewed in 2005 with the consequent reissue of the standard as ASTM F 838-05.

Application to integrity testing

Integrity testing of filters is one of a series of interdependent activities for the preparation of sterile drugs. The key requirement of an integrity test is to establish a quantitative correlation between a property or characteristic of the filter of interest and its organism retention capabilities. This would, for example, address the needs of sterilizing filtration practitioners to identify filters of a particular pore size rating that have the necessary bacterial retention capability.

FDA has defined a sterilizing filter as one that retains the classic challenge of at least 1×107 colony forming units of B. diminuta ATCC-19146/cm2 of effective filtration area at pressures up to 30 psi (2 bar). This is best assessed by direct confrontation of the membrane by a proper organism challenge. However, once the filter has been tested it is contaminated by the organisms and, therefore, may not be employed as a process filter. Therefore, a surrogate test is used that is nondestructive to filters, and is based on correlation of pore size with organism retention.

The recognition of a pertinent correlation involves more than the parallelling of two events; the relationship of the two actions that are seemingly correlated must accord to a technical logic as being the resultants of a common origin. The classical integrity test value or number correlates a filter's pore structure to its ability to completely retain challenges of a stipulated number of organisms per filter area. The significance of the integrity test is that it relates to its performance function.

In the present practice, the membrane's largest pore size is probed by the bubble point test. Its value can be correlated with the membrane's ability to sustain the 1×107 /cm2B. diminuta confrontation made by the use of the standard ASTM F 838-05 microbiological challenge procedure. On this basis, such membranes are experimentally qualified to be of the 'sterilizing grade,' and are, thus, labelled as being of the 0.2 (0.22) μ-rated pore size. They may then be selected for trials in actual product filtration sterilizations to ascertain their suitability to perform in given filtration contexts.

Filter manufacturers have measured and correlated bubble point/diffusive flow values and LRVs for commercially available membranes. It is this correlation that serves as a nondestructive test for determining a given filter's organism retention capabilities. It is ASTM standard F 838-05 that makes possible the industry wide definition of LRV that is essential to this correlation. This is the information required to ensure the proper choice of membrane for use in filtrative sterilizations.

Present implications

Because of ASTM standard F 838-05, filtration with 0.2 (0.22) μ-rated filters became dependable instead of a gamble as to whether the so-rated filter was what it claimed. The FDA Aseptic Guidance document of 1987 was the first to define a sterilizing grade filter based on the ASTM standard description.15 The ASTM standard has also been adopted by many different agencies; for example, ISO's 13408-2 and PDA Technical Report 26.16,17

This ASTM standard correlates the nondestructive integrity tests with the B. diminuta retention characteristics of specific pore sizes. It is employed in defining the test limits of sterilizing grade filters. FDA recommends all sterilizing grade membrane filters to be integrity tested before and after utilization. As the filter user is unable to directly challenge every single filter used with organisms, filter manufacturers developed nondestructive integrity test methods, such as bubble point or diffusive flow, to serve as surrogates. However, these integrity tests must be correlated to the ASTM F 838-05 organism challenge test to determine their compliance with maximum allowable diffusive flow or minimum allowable bubble point limits (Figure 2). Such correlation tests are performed by filter manufacturers within their development laboratories and are published in validation or instruction documents.

Figure 2

Furthermore, the standard became a cornerstone for the development of new filters in the 0.2 (0.22) μ-rated product area. Every time filtration products are improved or redesigned, these filters undergo standard bacteria challenge tests to verify the desired pore size rating. Moreover, while the ASTM F 838-05 standard defines the challenge test for 0.2 (0.22) μ-rated membrane filters, its guidance is of such value for filter manufacturers and users that its challenge conditions are also employed for other pore size ratings, except with different challenge organisms; for example, 0.45 μ-rated filters are challenged under ASTM test conditions with Serratia marcescens being used as the challenge organism. For 0.1 μ-rated filters the commonly used organism is Acholeplasma laidlawii. New standards for challenge tests of different pore sizes could, potentially, be based on ASTM F 838-05, as the test method is already 'unofficially' used by many filter manufacturers to verify pore size ratings.

In 1995, FDA employed the ASTM standard to further enhance sterility assurance by filtration. Most challenge tests performed before that point were commonly carried out on a water or broth basis under the teaching of FDA's challenge standard. However, the agency required verification of the sterile filtration performance of the filter under actual process conditions governing the filter's use and, if possible, with the actual drug product being the challenge fluid. The challenge level and the organism type stayed the same, and the challenge conditions and fluid stream utilized were adapted to production process reality. Again, this requirement has been prescribed by multiple organizations and guidances. A specific decision matrix has been established to define which actions should be taken when the challenge fluid is incompatible with the challenge organism (Figure 3).

Figure 3

Sterilizing grade and pore size rating

So successful were the 0.2 (0.22) μ-rated membranes in producing sterile effluent that the identification of a filter's pore size rating by way of integrity testing was taken to ensure that its use would yield a sterile effluent. The guiding assumption, however, was dependent upon sieve retention or size exclusion being the organism retention mechanism and was based on B. diminuta serving as the model organism.

With time, it was found that the 0.2 (0.22) μ-rated membranes did not necessarily result in a sterile effluent. It would seem that the key correlation based on the ASTM developed standard was not acceptable.

Further investigation showed that the organisms that escaped capture by the 'sterilizing' 0.2 (0.22) μ-rated membranes had undergone size diminution following immersion in liquid vehicles of limited nutritional value. Far from proving inadequate, the correlation demonstrated its applicability for retaining smaller organisms by the use of membranes of lower pore size ratings, namely, the so-called 0.1 μ-rated filters.

Filtration practitioners' understanding that the sizes of organisms were capable of undergoing change was heightened by Leahy and Sullivan who found that the size of the B. diminuta microbes depended upon the method of cultivation.18 This emphasized the need to standardize the procedure whereby the very organisms used to confront the filter pores would repeatedly and dependably be cultivated to match in size. The operational protocol set forth in ASTM standard F 838-05 made this possible, eliminating the issue of organism size variability. The standard satisfied the need for organisms of sufficiently similar size to serve as standard size particles. Consequently, they were suitable for the assaying of their retention by pores of different sizes. This legitimized comparisons among competitive filters on technical grounds as the standard eliminated challenge test variability.

0.1 versus 0.2 (0.22) μ-rated pore size filters

To increase effectivity of the sterilizing filter, the substitution of the 0.1 for the 0.2 (0.22) μ-rated membrane was suggested. However, of the seven 0.1 μ-rated filters tested, only four were found efficacious. The more generalized use of 0.1 μ-rated filters was, thus, unwarranted.

Correlation of LRVs with the assaying of organism counts could not be applied to a pore size classification that lacked industry identification. A suitable test organism whereby the ASTM standard F 803-05 test could provide the necessary correlation had first to be agreed upon.

In the circumstances, the use of the tighter filters was limited by practical considerations such as lower rates of flow. The substitution of 0.1 for the 0.2 (0.22) μ-rated filters is recommended where organism shrinkage is possible — especially for filtrations of long duration. The use of the tighter membrane awaits its LRV correlation to the native bioburden organism colony counts of its effluents. The native microbe appropriate to the inquiry requires identification. The ASTM standard creates a basis to perform such investigations.

Technology and commercialism

The success of filtration sterilization applications encouraged competition between filter manufacturers that some believe bordered on marketing rather than technical considerations.

In the circumstances, the fact that ASTM standard F 838-05 existed empowered the pharmaceutical industry to secure scientific-based correlation data that muted commercial concerns. Beneficially, the ASTM standard devoted to technical matters, in quieting commercial considerations, attained a laudable goal. Society and the patient were the beneficiaries of an ASTM standard-setting effort.

Utility in validations

Bioburden studies, integrity testing and process validation are the building blocks of filtration sterilization exercises. The contributions of ASTM standard F 838-05 to bioburden studies and to filter integrity testing have been discussed, along with its influence on fulfiling validation requirements. It created a well-defined testing range and activity flow for the validation efforts essential for verification of the filtrative sterilization operation. Without this industry standardization on behalf of a specific and acceptable test range, individual efforts would be confined to limited goals. The basis for the enhanced challenge definitions, and for the adoption of a true process validation approach was ASTM F 838-83 (now 838-05).

Conclusion

The ASTM standard F 838-83 (now 838-05) finally shed light on the retentive performances of filters previously rated sterilizing grade or 0.2 (0.22) μ. The standard became the foundation of many regulatory guidance documents and minimized the confusion regarding microbial retentivity. Although the standard was first published in 1983, it still poses a basis for multiple process validation approaches and has the potential for further creating standards regarding other retention ratings.

Ultimately, the ASTM F 838-83 (now 838-05) standard enhanced sterility assurance and, moreover, patient wellbeing.

Acknowledgement

The authors would like to express their gratitude and appreciation to everybody involved in the definition and establishment of ASTM F 838-83.

Maik Jornitz is Vice President Product Management FT/FRT at Sartorius Stedim North America Inc. (NY, USA).

Theodore H. Meltzer is Principle of Capitola Consultancy (MD, USA).

References

1. R. Zsigmondy and W. Bachman, Ueber Neue Filter. Zeitschrift Anorganische Chemie. 103:119, J. Soc. Chem. Ind. 37:453A (1918).

2. HIMA — Microbial Evaluation of Filters for Sterilizing Liquids, Document No 3, Vol 4, 1982.

3. ASTM, Committee F 838-83 — Standard Test Method for Determining Bacterial Retention of Membrane Filters Utilized For Liquid Filtration, 1983. www.astm.org

4. ASTM, Committee F 838-05 — Standard Test Method for Determining Bacterial Retention of Membrane Filters Utilized for Liquid Filtration, 2005. www.astm.org

5. F. Erbe, Kolloid Zeitung, 63(3), 277–285 (1933).

6. M.H. Alkan and M.J. Groves, Drug Dev. Ind. Pharm., 4(3), 225–241 (1978).

7. H. Yasuda and J.T. Tsai, J. Appl. Polym. Sci., 18, 805–819 (1974).

8. ASTM, Committee F-316 — Pore-Size Characteristics of Membrane Filters for Use with Aerospace Fluids, 1980. www.astm.org

9. F.W. Washburn, Proc. Natl. Acad. Sci. USA, 7, 115–116 (1921).

10. E. Honold and E.L. Skau, Science, 120, 805–806 (1954).

11. M.G. Katz, Measurement of pore-size distribution in microporous filters and membranes, Proceedings of World Filtration Congress III (Downington, PA, USA, 1982) pp 508–512.

12. P.R. Johnston, J. Testing Eval., 11(2), 117–125 (1983).

13. C.T. Badenhop, "The determination of the pore distribution and the consideration of methods leading to the prediction of retention characteristics of membrane filters," Doctoral Thesis, University of Dortmund, Germany (1983).

14. F.W. Bowmann, M.P. Calhoun and M. White, J. Pharm. Sci., 56(2), 453–459 (1967).

15. FDA Guidance for Industry — Sterile Drug Products Produced by Aseptic Processing, 1987 and 2004. www.fda.gov

16. ISO — Aseptic processing of healthcare products — Part 2: Filtration, 2003. www.iso.org

17. PDA — Technical Report 26, Sterilizing Filtration of Liquids, 1998. www.pda.org

18. T.J. Leahy and M.J. Sullivan, Pharma. Technol., 2(11), 64–75 (1978).