• A pure culture of the diminutive bacteria.
• No reversion to larger sizes.
• Sufficient quantities for bacterial challenge studies.

Although the numerical ‘pore size’ definition helps us to classify filters into general groups such as 0.45µm, 0.2µm, 0.1µm, etc., a more meaningful definition must include some reference to its microbiological removal capability. In this way, the potential for specific bacterial types to be retained by the filter may be more accurately determined. This approach requires specification of the organism, the laboratory test method and a removal efficiency level if the definition is to become a meaningful and appropriate industry standard.

However, even a definition and specification based on removal efficiency does not necessarily guarantee sterility in the process. The filter may not be physically or chemically compatible with the process. Also, bacteria smaller than B. diminuta may be present. These diminutive forms may be naturally occurring or induced by process conditions. Process validation may be able to establish the most suitable sterilizing filter type and rating, but it is difficult to predict or monitor during routine production the type and quantity of bioburden present in every product batch. For this reason, there has been a major increase, especially over recent years, in the level of process validation required to justify the use of the current 0.2µm or 0.22µm sterilizing filters.

This publication reviews the evolution of sterile filtration with specific reference to penetration by specific organisms. New data are also presented on a diminutive organism, isolated from water, which can be grown in quantity under laboratory conditions without losing its diminutive form. This organism has been shown to penetrate various commercially available 0.2µm or 0.22µm sterilizing filters, but not 0.1µm filters. The implications for filter validation are discussed and the future role for 0.1µm filtration assessed.

In the 1960s, both sterilizing sheet filters and membrane filters were used in pharmaceutical production. Membrane filters were available mostly in flat disc form and used singly or in multi-plate configuration. 0.45µm membranes were successfully used for sterile production at that time as they enabled reasonable flow rates to be achieved through the relatively low area disc systems. The membranes were qualified using Serratia marcescens, with a typical size of 0.6µm x 1µm. However, the safe use of 0.45µm filters was questioned when Bowman² established that an organism, Pseudomonas diminuta, could consistently penetrate 0.45µm ‘sterilizing’ filters, but could be retained by the next finer grade commercially available - 0.22µm.

0.2µm/0.22µm filter standard.

Bowman2 proposed in 1967 that P. diminuta (recently reclassified as Brevundimonas diminuta) should become the industry standard organism for 0.2µm filters. In 1987, the FDA ‘Guidelines on sterile drug products produced by aseptic processing’³ incorporated P. diminuta as the standard challenge organism for a sterilizing filter and defined a minimum qualifying level of 10⁷/cm² of filter area.

Since that time, no further standards have been developed, even though 0.1µm sterilizing filters in cartridge form have been available commercially for almost twenty years and are being increasingly used in production processes. The primary area of application was initially in the processing of serum and tissue culture media, where removal of mycoplasma is required. These deformable bacteria are known to penetrate 0.2µm or 0.22µm filters. However, there has been an increasing use of 0.1µm filters in other applications where diminutive organisms have been identified, or are of potential concern. They are also being used for enhanced sterility assurance in certain types of products or processes.

In the absence of a defined industry standard, filter manufacturers have qualified 0.1µm filters using their own standards. PALL uses Acholeplasma laidlawii, a mycoplasma type organism, for 0.1µm filter validation⁴ in addition to B. diminuta.

• Substantial and rapid penetration of all 0.2µm and 0.22µm filters by Pseudomonas sp. giving downstream counts of between 10² and 10⁵ cfu.
• Full retention of B. diminuta.
• Total removal of both Pseudomonas sp. and B. diminuta by 0.1µm filters.

Further work is in progress to identify the diminutive species and to assess its suitability for wider use in filter qualification testing.

Table III

Mixed bacterial challenge of sterilizing filter cartridges *
Filter Type	Rating	B. diminuta		Pseudomonas sp.		Sterile
		Challenge	Recovery	Challenge	Recovery
PALL Ultipor® N66® Grade NF/NR	0.2µm	1 x 1011 1 x 1011 9 x 1010	0 0 0	2 x 1010 6 x 109 2 x 109	1 x 105 1 x 104 1 x 104	No No No
PALL Fluorodyne® II Grade DFL	0.2µm	1 x 1011 7 x 1010 1 x 1011	0 0 0	1 x 1010 3 x 109 4 x 1010	1 x 104 2 x 102 7 x 102	No No No
Non-PALL PVDF	0.22µm	1 x 1011 8 x 1010 8 x 1010	0 0 0	1 x 109 9 x 109 8 x 108	1 x 104 1 x 104 1 x 104	No No No
PALL Ultipor® N66® Grade NT	0.1µm	1 x 1011 1 x 1011 6 x 1010	0 0 0	2 x 109 3 x 109 4 x 1010	0 0 0	Yes Yes Yes
PALL Fluorodyne® II Grade DJL	0.1µm	1 x 1011 1 x 1011 2 x 1011	0 0 0	6 x 109 3 x 109 5 x 1010	0 0 0	Yes Yes Yes

*All studies used 254mm (10 inch) cartridges. Flow rate 0.3 L/min.

The same organisms were fully retained by a PALL Ultipor^® N₆₆^® (grade NT) 0.1µm rated filter, as shown in Table I.

Table I

Removal of diminutive bacteria by 0.1µm filters7
Filter Number	Filter Size	Total Challenge	Bacteria Recovered
1	142mm Disc	1.2 x 107	0
2	293mm Disc	8.9 x 107	0
3	254mm Cartridge	1.7 x 1010	0

• In 1985, Andersen et al.⁸ reported penetration of 0.2µm filters by Burkholderia (Pseudomonas) pickettii in saline solution.
• In 1993, the FDA reported presence of Pseudomonas cepacia in deionized water downstream of 0.2µm filters⁹. In the same document, they reported a product recall in the USA for a solution of povidone iodine which was also suspected to involve penetration of the 0.2µm filter by P. cepacia.
• In a recent publication¹⁰, Leo et al. reported penetration of 0.2µm and 0.22µm filters by Burkholderia (Pseudomonas) pickettii when suspended in product, but not when suspended in saline lactose broth (SLB). The penetration was associated with a 40% reduction in bacterial size, as shown in Figure 3. Full retention was achieved using a PALL 0.1µm Ultipor^® N₆₆^® filter. Further work is in progress to determine the mechanism of filter penetration.

Figure 3
Size distribution of B. pickettii suspended in product¹⁰

Events of this type have resulted in an increasing requirement by regulatory authorities for product and process specific filter validation. Such additional validation is required to be performed in a way that simulates as closely as possible the actual production conditions and also represents ‘worst case’³.

More recently, the need to challenge test in the actual drug product being filtered has been emphasised by the FDA¹¹, who have stated: “Since there is the possibility that the drug product may cause a reduction in the size of the micro-organism, it is best to test the microbial retentivity of the filter with the microbial challenge in the actual drug product.” and “…There are products that fall within Millipore’s matrix which are not rendered sterile when filtered through a 0.22 micron filter.”

Despite increased validation for 0.2µm and 0.22µm filters, the FDA¹² in 1996 have commented that the use of a 0.1µm filter as the final filter may be “desirable to ensure sterility of the product, especially when any contaminants are exposed to the drug product over a long period of time.”

1. M.Osumi, N. Yamada and M. Toya. “Bacterial retention mechanisms of membrane filters”. PDA Journal of Pharmaceutical Science and Technology 50: 30-34, (1996)
   
2. F.Bowman, M.P. Calhoun and M. White, “Microbiological methods for quality control of membrane filters”, J. Pharm. Sci., 55, 818 (1967)
   
3. FDA, “Guidelines on sterile drug products produced by aseptic processing”. Centre for Drugs and Biologics, Rockville. MD, (1987)
   
4. “Validation guide for Pall NT grade 0.1µm microbially-rated nylon 66 membrane cartridges”. Pall Corporation, East Hills, NY
   
5. “Rules governing medicinal products in the European Community Vol IV - Good manufacturing practice for medicinal products”. ISBN 92-826-3180-X. Office for official publications of the European Communities, Luxembourg, (1992)
   
6. J.L. Braun, S.L. Diesch and W.F McCulloch “A method for isolating leptospires from natural surface waters”. Canadian Journal of Microbiology 14: 1011-1012 (1968).
   
7. G. Howard and R. Duberstein. “A case of penetration of 0.2µm rated membrane filters by bacteria.” Journal of the Parenteral Drug Association.,34: p95 (1980)
   
8. R.L. Anderson, L.A.Bland, M.S. Favero, M.M. McNeil, B.J. Davis, D.C. Mackel and C.R. Gravelle. “Factors associated with Pseudomonas pickettii intrinsic contamination of commercial respiratory therapy solutions marketed as sterile”. Applied and Environmental Microbiology 58: 1343-1348. (1985)
   
9. D.L. Michels “Validation and control of de-ionised water systems”. FDA, Rockville MD20857, (1993)
   
10. F. Leo, M. Auriemma, P. Ball and S. Sundaram, “Application of 0.1µm filtration for enhanced sterility assurance in pharmaceutical filling operations”. BFS News, August (1997) (in press)
    
11. Human Drug CGMP Notes. Vol 2, No. 4, p35. FDA CDER Office of Compliance. (1995)
    
12. CDER Perspective on Isolator Technology ISPE Barrier Isolation Technology Conference, Rockville MD (1996)