Filtration is a common method of sterilizing drug product solutions. An appropriate sterilizing grade filter is one which reproducibly removes all microorganisms from the process stream, producing a sterile effluent. Such filters usually have a rated porosity of 0.2 micron or smaller.
Whatever filter or combination of filters is used, validation should include microbiological challenges to simulate “worst case” production conditions regarding the size of microorganisms in the material to be filtered and integrity test results of the filters used for the study. The microorganisms should be small enough to both challenge the nominal porosity of the filter and simulate the smallest microorganism that may occur in production. The microorganism Brevundimonas diminuta (ATCC 19146) when properly grown, harvested and used, can be satisfactory in this regard because it is one of the smallest bacteria (0.3 micron mean diameter).
Bioburden of unsterilized bulk solutions should be determined, in order to trend the characteristics of potentially contaminating organisms. In certain cases, when justified as equivalent or better than use of Brevundimonas diminuta, it may be appropriate to conduct bacterial retention studies with a bioburden isolate. The number of microorganisms in the challenge is important because a filter can contain a number of pores larger than the nominal rating which have potential to allow passage of microorganisms. The probability of such passage is considered to increase as the number of organisms (bioburden) in the material to be filtered increases. A challenge concentration of at least 107 organisms per cm2 of effective filtration area of B. diminuta is generally used. A commercial lot’s actual influent bioburden should not include microorganisms of a size and/or concentration that would present a challenge beyond that considered by the validation study.
Direct inoculation into the drug formulation provides an assessment of the effect of drug product on the filter matrix and on the challenge organism. However, directly inoculating B. diminuta into products with inherent bactericidal activity or into oil-based formulations can lead to erroneous conclusions. When sufficiently justified, the effects of the product formulation on the membrane’s integrity can be assessed using an appropriate alternate method. For example, the drug product could be filtered in a manner in which the worst-case combination of process specifications and conditions are simulated. This step could be followed by filtration of the challenge organism for a significant period of time, under the same conditions, using an appropriately modified product (e.g., lacking an antimicrobial preservative or other antimicrobial component) as the vehicle. Any divergence from a simulation using the actual product and conditions of processing should be justified. Factors which can affect filter performance normally include:
(1) viscosity of the material to be filtered; (2) pH; (3) compatibility of the material or formulation components with the filter itself; (4) pressures; (5) flow rates; (6) maximum use time; (7) temperature; (8) osmolality; (9) and the effects of hydraulic shock.
When designing the validation protocol, it is important to address the effect of the extremes of processing factors on the filter capability to produce sterile effluent. Filter validation should be conducted using the worst case conditions, such as maximum filter use time and pressure. Filter validation experiments, including microbial challenges, need not be conducted in the actual manufacturing areas. However, it is essential that laboratory experiments simulate actual production conditions. The specific type of filter used in commercial production should be evaluated in filter validation studies. When the more complex filter validation tests go beyond the capabilities of the filter user, tests are often conducted by outside laboratories or by filter manufacturers. However, it is the responsibility of the filter user to review the validation data on the efficacy of the filter in producing a sterile effluent. The data should be applicable to the user’s products and conditions of use because filter performance may differ significantly for various conditions and products.
After a filtration process is properly validated for a given product, process and filter, it is important to ensure that identical filter replacements (membrane or cartridge) used in production runs will perform in the same manner. Sterilizing filters should be routinely discarded after processing of a single batch. Normally, integrity testing of the filter is performed after the filter unit is assembled and sterilized prior to use. It is important that the integrity testing be conducted after filtration in order to detect any filter leaks or perforations that might have occurred during the filtration. “Forward flow” and “bubble point” tests, when appropriately employed, are two acceptable integrity tests. A production filter’s integrity test specification should be consistent with data generated during filtration efficacy studies.
