Method for qualifying microbial removal performance of 0.1 micron rated filters. Part III: bacterial challenge tests on 0.2/0.22 and 0.1 micron rated filter cartridges with Hydrogenophaga (formerly Pseudomonas) pseudoflava.

We have previously reported on the preliminary characterization of Hydrogenophaga (formerly Pseudomonas) pseudoflava for potential use as a standard challenge organism to qualify 0.1 microm rated filters. This article reports on the retention efficiencies of a large panel of 0.2/0.22 microm and 0.1 microm rated filter cartridges for H. pseudoflava (ATCC 700892) versus the retention capabilities of the same filters for Brevundimonas diminuta (ATCC 19146). A total of thirty-two 0.2/0.22 microm rated filter cartridges, spanning nine different "sterilizing grade" filter types from four different filter manufacturers, were challenged with H. pseudoflava at challenge levels exceeding 10(7) cfu/cm2. H. pseudoflava was shown to penetrate every 0.2/0.22 microm rated filter tested, with log titer reduction (LTR) values ranging from 3.5 to 7.7 logs. H. pseudoflava was shown to be more penetrative than B. diminuta under the same challenge conditions. B. diminuta was fully retained by nineteen of the twenty 0.2/0.22 microm rated filters that were challenged with both organisms. In the case of 0.1 microm rated filters, eighteen filter cartridges, spanning five different filter types from three manufacturers were tested. H. pseudoflava was consistently retained by four out of the five filter types tested, with LTR values in excess of 11.5 to 12.2 logs. The 0.1 microm rated filter type that was penetrated by H. pseudoflava has been previously demonstrated to be not fully retentive for naturally occurring bacteria. The data show that H. pseudoflava penetrates 0.2/0.22 microm rated filters just as readily as B. diminuta penetrates 0.45 microm rated filters. In addition, titer reductions provided by 0.2/0.22 microm rated filters for H. pseudoflava are comparable to those reported for A. laidlawii mycoplasma, albeit under different conditions. This study demonstrates that H. pseudoflava meets all criteria for use as a standard organism for qualifying the microbial removal performance of 0.1 microm rated filters for enhanced sterility assurance.     

Method for qualifying microbial removal performance of 0.1 micron rated filters. Part I: characterization of water isolates for potential use as standard challenge organisms to qualify 0.1 micron rated filters.

Although 0.1 microm rated filters intended for pharmaceutical sterilization applications have been commercially available for at least 15 years, there is no industry-wide standard for qualifying the microbial removal performance of these filters. In this article, we report on the bacterial challenge methodology used to screen four bacterial species for potential utility as a standard challenge organism to qualify 0.1 microm rated filters. These isolates were, in their natural state, demonstrated to penetrate 0.2/0.22 microm rated filters in prior studies. In the screening challenges described in this study, three out of these four candidates tested demonstrated consistent penetration of one 0.22 microm rated filter type tested (when cultured in a low nutrient medium under standard laboratory conditions). These included 6204-22 (FAME ID Acidovorax avenae citrulli), 6266-15 (FAME ID Comamonas acidovorans), and 6266-34 (FAME ID Hydrogenophaga pseudoflava). Of these, H. pseudoflava (6266-34) was chosen for additional experiments with other 0.2 microm rated filter membranes. In total, seventeen 0.2 and 0.22 microm rated filter discs, spanning five different "sterilizing grade" filter types from three different filter manufacturers were tested. H. pseudoflava penetration was observed for every filter tested. Under the same challenge conditions, H. pseudoflava was consistently retained by a 0.1 microm rated hydrophilic PVDF (polyvinylidenefluoride) filter with a specified high titer reduction claim for Acholeplasma laidlawii. In order to ensure selection of the most stable penetrative phenotype (i.e., select for nonrevertants), H. pseudoflava was subjected to three rounds of "filter cloning," and these results are described herein. The advantages of using H. pseudoflava for qualifying the microbial removal performance of 0.1 microm rated filters are also discussed.     

Method for qualifying microbial removal performance of 0.1 micron rated filters. Part IV: Retention of hydrogenophaga pseudoflava (ATCC 700892) and Ralstonia pickettii (ATCC 700591) by 0.2 and 0.22 micron rated filters

Ralstonia pickettii has emerged as a bioburden microorganism of considerable importance in pharmaceutical processes utilizing conventional 0.2 or 0.22 micron rated "sterilizing grade" filters. In this article, we re-evaluated and studied the retention efficiencies of 0.2 micron rated nylon 6.6 and 0.22 microns rated modified polyvinylidene fluoride (PVDF) filters for Hydrogenophaga pseudoflava (ATCC 700892) and R. pickettii (ATCC 700591). Out of a total of forty-four 0.2/0.22 micron rated filters discs tested in this study (spanning different challenge fluids, different challenge conditions, and different filter types), H. pseudoflava penetration was observed for every filter disc tested. Log titer reduction (LTR) values ranged from 0.3 to 2.0 logs for 20-48 hour challenges conducted in Water for Injection (WFI), and 3.8-7.1 logs for 6-hour challenges conducted in Minimal Media Davis (MMD). For 0.2 micron nylon 6.6 filter discs, penetration by R. pickettii was observed only in WFI challenges and was dependent on the culture and challenge conditions used. Penetration by R. pickettii was also restricted to only those membrane discs that were very close to the filter manufacturer's production integrity test (the Quantitative Bubble Point, QBP, test) limit. Where R. pickettii penetration was observed, LTR values were significantly higher than those observed for H. pseudoflava with the same filter discs. This study: 1) supports the use of H. pseudoflava as a worst-case challenge model for R. pickettii in process- and product-specific bacterial retention testing; 2) provides experimental evidence, for the first time, for the need to include filter membrane lots that have a physical integrity test value at or near the filter manufacturer's production (lower) limit in these tests; and 3) demonstrates how a standardized membrane integrity test (such as the QBP test) can be used select such "worst-case" membranes and to verify the inclusion of such "worst-case" membranes in these tests, thus serving as the link between the membrane disc used in bacterial retention validation testing and the production process filter.     

Retention of water-borne bacteria by membrane filters. Part III: Bacterial challenge tests on 0.1 micron rated filters

Clear performance differences were observed between different 0.1 micron rated filters in terms of their microbial removal efficiency when challenged with naturally occurring waterborne bacteria from a water source. Penetration occurred with three 0.1 micron rated "sterilizing grade" filter types tested, from three different filter manufacturers, that did not have a specific high titer reduction claim for Acholeplasma laidlawii. Bacteria shown to penetrate these 0.1 micron rated filters were quite similar to those recovered downstream of 0.2.0.22 micron rated filters (described in Part II). All of the isolates identified via FAME analyses were common environmental or ubiquitous organisms, and some, such as Acidovorax sp. and Hydrogenophaga pseudoflava, have also been isolated from pharmaceutical water systems. In contrast, four different 0.1 micron rated "sterilizing grade" filter types from two different manufacturers, which had been qualified with both B. diminuta and A. laidlawii, consistently produced sterile effluents under similar test conditions. This study thus highlights the need for an industry or regulatory standard method of defining the microbial removal performance of 0.1 micron rated filters, and supports the use of functionally qualified 0.1 micron rated filters as sterilizing grade filters in pharmaceutical operations for enhanced sterility assurance.     

Retention of water-borne bacteria by membrane filters. Part I: Bacterial challenge tests on 0.2 and 0.22 micron rated filters

The results of bacterial challenge tests conducted on several 0.2 and 0.22 micron rated "sterilizing grade" filter cartridge types with bacteria from a natural water source are presented. Eight different 0.2/0.22 micron rated "sterilizing grade" filter types from four different filter manufacturers, claimed to be capable of retaining Brevundimonas diminuta at a challenge level of 10(7) CFU/cm2, were tested. The filters tested included nylon 6.6 and polyamide filters from two manufacturers, modified or hydrophilic PVDF filters from two manufacturers, modified or asymmetric PES filters from three manufacturers, and cellulose acetate filters from a single manufacturer. Consistent bacterial penetration was observed, over the 18-24 h challenge period, for all twenty-five integral 0.2 and 0.22 micron rated filter cartridges tested, at challenge levels of about 10(1)-10(4) CFU/cm2, indicating that natural waterborne bacteria were more penetrative than B. diminuta. The observed penetration was thus qualitatively independent of filter media type or manufacturer. These results add to the growing body of evidence that shows 0.2 and 0.22 micron rated filters may not remove all microorganisms under all conditions. These results further establish that bacterial penetration of 0.2/0.22 micron rated filters is not limited just to (1) specific membrane types, or (2) extended duration challenges (> 24 h), or (3) extremely high challenge levels, or (4) bacteria that can only exist in a penetrative state in an artificial laboratory setting.     

Retention of water-borne bacteria by membrane filters. Part II: Scanning electron microscopy (SEM) and fatty acid methyl ester (FAME) characterization of bacterial species recovered downstream of 0.2/0.22 micron rated filters

The results of scanning electron microscopic (SEM) and fatty acid methyl ester (FAME) characterization of the bacterial species shown to penetrate conventional 0.2/0.22 micron rated "sterilizing grade" filters are presented. SEM data suggest that retention of bacteria by these filters appears to be strongly influenced by the morphology, and especially the width of bacteria and less so by length. When the bacterial cell width is small, less than 0.3 micron or so, the cell length does not appear to limit the ability to penetrate 0.2/0.22 micron rated filters. As the bacterial width increases, there is also a strong, almost exponential, decrease in the allowable length for penetration, with most penetrative cells tending to be coccoid beyond a width of 0.5 micron. Significant percentages of the bacteria (40-50%) that were observed downstream of these filters were larger than B. diminuta, the standard organism used to qualify 0.2/0.22 micron rated filters. The average sizes of natural waterborne bacteria that penetrated the filters tested were 20-40% larger in width, and 40-70% larger in length, compared to B. diminuta. These results indicate that size exclusion is not the sole mechanism governing bacterial retention. All isolates identified via FAME analyses were common environmental or ubiquitous organisms, and some, such as Acidovorax sp. and Hydrogenophaga pseudoflava, have also been isolated from pharmaceutical water systems. Most of the bacteria recovered downstream of 0.2/0.22 micron rated filters were gram negative, oxidase positive, motile, nonfermentors.     

Application of membrane filtration for removal of diminutive bioburden organisms in pharmaceutical products and processes

In this report, we present results of a recent investigation in our laboratories demonstrated the effect of process conditions and/or drug product composition on the ability of 0.2 micron and 0.22 micron sterilizing grade filters to fully retain Ralstonia (formerly Burkholderia, formerly Pseudomonas) pickettii. R. pickettii is a opportunistic pathogen widely distributed in nature as well as clinical specimens and there have been several reports of nosocomial infections due to intrinsic manufacture-related R. pickettii contamination in filter-sterilized parenteral fluids. This study documents the penetration of 0.2 micron nylon 66 and 0.22 micron modified PVDF sterilizing grade filters by R. pickettii (grown and challenged) in a drug solution under conditions that simulated a pharmaceutical filling operation. Penetration was not observed for every filter disc tested, and this may be explained, in part, by the stochastic nature (i.e., governed by the rules of probability) of the retention mechanisms involved. Scanning electron microscopy revealed significant changes in the microorganism's size and morphology as a result of exposure to the drug solution; these changes are consistent with those reported for bacteria subjected to nutrient deprivation. The SEM analyses of R. pickettii challenge suspensions in the drug solution showed that the average cell length decreased from 1.25 +/- 0.27 microns to 0.84 +/- 0.17 micron between zero and 24 hours. In addition, significant changes were observed in the size (length) distributions, with approximately 35% of the cells at 24 hours being smaller than any cell observed at the start of the challenge. These data suggest that the significant reduction in bioburden size and morphology that occurred as a result of exposure to the drug solution may play a role in the reduced ability of the 0.2 micron and 0.22 micron filters tested in this study to retain these organisms. Under the same test conditions where penetration of 0.2/0.22 micron filters was observed, 0.1 micron rated membrane filters qualified with both B. diminuta and Acholeplasma laidlawii mycoplasma consistently provided sterile effluent. Bacterial penetration of 0.2 (or 0.22) micron sterilizing grade filters was not observed under identical test conditions with either R. pickettii in a standardized solution (saline lactose broth) routinely used in challenge testing filters, or with the standard test organism, B. diminuta, in the drug solution. This study thus supports the renewed emphasis on both product- and process specific validation as well as routine bioburden monitoring expressed by regulatory agencies, and the use of enhanced bacterial removal efficiency 0.1 micron rated filters to provide enhanced sterility assurance in pharmaceutical processes.     

Effects of sterilizing-grade filters on the physico-chemical properties of onion-like vesicles

Spherulites are new promising multilamellar vesicles that we study in a drug delivery context. The sterilization of spherulites suspensions is a necessary step before biological tests and later, before pharmaceutical applications (for example, parenteral or local injections). Among all sterilizing operations, the filtration through 0.22 microm sterilizing-grade filters (of the type Millex (? 4 mm) by Millipore) is easy and rapid, and we decided to study it as a mean to obtain sterile suspensions. The spherulites diameter is usually comprised between 0.2 and 0.5 microm but bigger vesicles occur and reach ? 1 microm. The effects of such filters on vesicles' size and lipids' concentration were then compromised. After examination of this challenging operation, results proved that the sterilizing filtration had no effect on these two parameters whatever the formulation chosen. Then, the possible release of amaranth, an encapsulated hydrophilic dye was followed. With the formulations and in spite of a filter diameter inferior to that of the vesicles, the encapsulation yields were not significantly different before and after the filtration and no leakage could be detected. Finally, the spherulites' functionality after sterilizing filtration was studied under the chemical angle: vesicles containing an amphiphilic reactive anchor (CholE3ONH2) were still able to bind covalently a peptidic molecular recognition pattern. The ligation was quantified by fluorimetry as high as for non-filtrated suspensions. Thus, though spherulites can present a diameter superior to that of the sterilizing filters, their passage through them do not alter the physico-chemical properties of these vesicles.     

Membrane filters and membrane-filtration processes for health care

The development of membrane-filtration processes is reviewed, and current types and uses of membrane filtration in health care is discussed. Development of adequate support structures for filters and of disposable filtration devices has facilitated development of filtration processes for pharmaceutical industry, manufacturing in hospital pharmacies, and direct patient care. Hydrophobic filters have also been developed; aqueous solutions cannot wet the pore structures of these filters and therefore cannot pass. Sterility-testing systems have also been developed. There are two types of filters: depth (constructed of compacted fibers) and membrane (which have a homogeneous internal structure). Depth filters retain only a portion of particles in a particular size range and are generally not acceptable for use in health care. Membrane filters retain all particles of a given size. Types of membrane filters are selected for specific uses based on needed flow rates, particulate load, and retention capability. Membrane filters may be validated using bacterial-passage, bubble-point, and diffusion tests. Most membrane filters used in health care are microporous filters that retain particles in the 0.1-10-micron size range. Applications are currently being developed for ultrafilters, which retain both particles and substances with large molecular structures such as proteins, and reverse-osmosis filter membranes, which allow only water or water-miscible solvents of very low molecular weights to pass. Experience in engineering designs, quality assurance, and test procedures has led to the development of many safe, reliable, and effective membrane products for health care.     

In line final filters for removing particles from amphotericin B infusions.

The feasibility of using membrane filters to remove particles from intravenous infusions of amphotericin B in dextrose 5% (a colloidal solution) was studied. Six types of commercial membrane filters, ranging in pore size from 0.45-1.0 mum, were used. Because of the effect of pH on the particle size of colloidal solutions, each filter was tested at solution pH 4.7, 5.6 and 6.5. Samples of filtrate were analyzed spectrophotometrically for amphotericin content. All filters of pore size 0.22 mum removed amphotericin B from solution and were inappropriate for use with this product. Solutions at pH 4.7 were turbid, filtered slowly and were generally unacceptable for clinical use. At pH 5.6, only filters with pore sizes of 1.0 mum or greater showed acceptable results. At pH 6.5, filters with pore sizes of 0.45 mum or greater gave acceptable results; the use of a filter with a pore size of not less than 1.0 mum would provide a margin for error to compensate for variations in the colloidal particle size of amphotericin B.     

Characterization of filter extractables by proton NMR spectroscopy: studies on intact filters with process buffers

Studies were conducted to characterize potential extractables from sterilizing grade filters. The focus of this report is the 0.22 micron Durapore (hydrophilic modified PVDF) filter which is used throughout our recovery processes. The objectives of this study are (1) to identify potential filter extractables from the hydrophilic PVDF filters; (2) to show that NMR spectroscopy may be used to detect filter extractables in the presence of product and excipients; and (3) to establish levels of filter extractables obtained by extraction with a variety of buffers. The data show that the primary source of filter extractables is the hydrophilic modification of the PVDF membrane surface. Extractables from the modified hydrophilic PVDF filter include propylene glycol (PG) and soluble oligomers of the hydroxypropyl acrylate and cross-linker. Propylene glycol, arising from the hydrolysis of the hydroxypropyl acrylate, appears to be the primary extractable in buffers above pH 11. Since the 1H-NMR method can easily detect the methyl proton signals of PG, an NMR assay was developed to detect PG in the presence of buffer excipients and final product. Propylene glycol can be used as a marker for the extractables from Durapore hydrophilic PVDF filters. Although numerous buffers were used to generate extractables from the PVDF filter, significant extractables (PG and soluble oligomers) were found only in high pH extraction buffers. As a result of this finding, only a limited number of new buffers or new PVDF filters will require testing for future validation studies. Process validation studies have shown that neither PG nor soluble oligomers are at levels that impact the quality or safety of the product.     

Impact of tubing material on the failure of product-specific bubble points of sterilizing-grade filters

The following study was conducted to determine the effect of different preservatives commonly used in the biopharmaceutical industry on the product-specific bubble point of sterilizing-grade filters when used to filter product processed with different types of tubing. The preservatives tested were 0.25% phenol, m-cresol, and benzyl alcohol. The tubing tested was Sani-Pure (platinum-cured silicone tubing), Versilic (peroxide-cured silicone tubing), C-Flex, Pharmed, and Cole-Parmer (BioPharm silicone tubing). The product-specific bubble point values of sterilizing grade filters were measured after the recirculation of product through the filter and tubing of different types of materials for a total contact time of 15 h. When silicone tubing was used, the post-recirculation product-specific bubble point was suppressed on average 13 psig when compared to the pre- recirculation product-specific bubble point. Suppression was also observed with C-Flex, but to a much lesser extent than with silicone tubing. Suppression was not observed with Pharmed or BioPharm tubing. Alcohol extractions performed on the filters that experienced suppressed bubble points followed by Fourier transform infrared spectroscopy analysis indicated the filters contained poly(dimethylsiloxane). Direct addition of poly(dimethlysiloxane) to solutions filtered through sterilizing-grade filters suppressed the filter bubble points when tested for integrity. Silicone oils most likely reduced the surface tension of the pores in the membrane, resulting in the ability of air (or nitrogen) to pass more freely through the membrane, causing suppressed bubble point test values. The results of these studies indicate that product-specific bubble point of a filter determined with only product may not reflect the true bubble point for preservative-containing products that are recirculated or contacted with certain tubing for 15 h or greater. In addition, tubing material placed in contact with products containing preservatives should be evaluated for impact to the product-specific bubble point when being utilized with sterilizing-grade filters.     

Experimental determination of the diffusion boundary layer width of micron and submicron particles

Powder dissolution kinetics have shown that for particles in the so called "large" size regime (more than about 50 microm), the dissolution rate scales as the specific surface area, i.e. rate proportional to d(-1) where d is the particle diameter. This is consistent with an effective diffusion boundary layer width h(EFF) that is constant with respect to particle size. However, for particles in the so called "small" size regime (d less than about 50 microm), the dissolution rate has a stronger dependence than proportional to d(-1) [Bisrat, M., Anderberg, E.K., Barnett, M.I., Nystroem, C., 1992. Physicochemical aspects of drug release. XV. Investigation of diffusional transport in dissolution of suspended, sparingly soluble drugs. Int. J. Pharm., 80, 191-201; Mosharraf, M., Nystroem, C., 1995. The effect of particle size and shape on the surface specific dissolution rate of microsized practically insoluble drugs. Int. J. Pharm., 122, 35-47]. In this regime, Prandtl boundary layer theory predicts an h(EFF) approximately equal to the particle radius or diameter. This paper presents the first experimental determination of h(EFF) for particles less than about 2 microm. The powder dissolution kinetics of six suspensions over the particle diameter range of 5.9 +/- 0.1 to 0.53 +/- 0.05 microm are analyzed to yield h(EFF) values of 8.5 +/- 1.9 to 0.34 +/- 0.14 microm. The theoretical expectation for mass transport, dissolution time proportional to d(2.0), is in good agreement with the experimental results of dissolution time proportional to d(2.3). An understanding of these mass transfer mechanisms allows pharmaceutical scientists to achieve targeted release rates with minimum ensemble instability.     

DNA solution(R) in cigarette filters reduces polycyclic aromatic hydrocarbon (PAH) levels in mainstream tobacco smoke.

Tobacco consumption represents a major health hazard to humans and, despite anti-smoking campaigns, the number of smokers remains high; thus the reduction of toxic compounds from tobacco smoke may reduce the health hazards of smoking. In the last 25 years cigarette manufacturers have introduced a variety of filter designs to reduce toxic and carcinogenic substances in tobacco smoke (normal filters, NF). However, large quantities of harmful constituents are inefficiently retained by commonly used cigarette filters. Following a patented method we modified commercial cigarette filters (modified filter, MF) by injecting a DNA solution into the filter tips; we then evaluated the reduced polycyclic aromatic hydrocarbon (PAH) levels in mainstream tobacco smoke of MF relative to NF. The PAH measured were: fluoranthene (FLUO), pyrene (PY), benzo(a)anthracene (B(a)A), chrysene (CRY), benzo(a)pyrene (B(a)P), benzo(b)fluoranthene (B(b)F), benzo(k)fluoranthene (B(k)F), benzo(g,h,i)perylene (BGP), dibenzo(a,h)anthracene (DBA). The levels of PAH in cigarette smoke after MF were significantly reduced (P<0.001) compared to NF, using a variety of cigarette brands in a smoking machine (44.5%+/-8.4 % and 41.8%+/-5% for total and carcinogenic PAH, respectively, means+/-SE). Using B(a)P(TEF) values the reduction in PAH concentrations were similar for all cigarette brands with the exception of Camel, where the reduction was lower considering B(a)P(TEF) values. Amongst carcinogenic PAH, B(a)A, B(b)F and B(k)F) were reduced by 50-58%, CRY, B(a)P and DBA by about 40%. In conclusion MF filters treated with DNA have the potential of decreasing the exposure to PAH in cigarette smoke. Since, unlike some previously proposed biological filters MF do not retain additional nicotine, the main addictive compound of tobacco smoke, these filters may not induce increased smoking to compensate for the reduction in the nicotine delivery to smokers.     

Changes in the cell size of Brevundimonas diminuta using different growth agitation rates

Brevundimonas diminuta (ATCC 19146) is a standard organism for validation of sterilizing-grade membrane filters. Cell size is critical for the determination of retention characteristics of 0.2 micron rated membrane filters. In this study, cell size changes of B. diminuta cultured under different physiologic states and variable agitations at 50, 100 and 200 rpm were measured by a particle size analyzer and scanning electron microscope (SEM). The smallest cells were obtained at initial stationary phase in saline lactose broth (SLB) as a shaking culture at 50 rpm. Cells grown under agitation at 50, 100 and 200 rpm showed an increase of specific growth rate (mu), about 2.9, 3.6 and 3.6 fold, respectively, compared to the non-agitated cells in SLB media. These results suggested that the cell size decreased proportionally with increase of the specific growth rate (mu) in SLB. These size changes were associated with penetration through a 0.2 micron rated cellulose acetate filter. A scale-down filtration system was developed and performed bacterial challenge test and bubble point test with cells cultured in SLB. Cells grown under agitation conditions in SLB were not retained by 0.2 micron rated membrane filter.     

Sepsis in rabbits following administration of contaminated infusions through filters of various pore sizes

The effects of three final filter pore sizes on sepsis and survival times in rabbits infused with contaiminated intravenous fluids were compared. Intravenous fluids were administered for up to 12 hours to five groups of rabbits. Groups 2, 3 and 4 received fluids, contaminated with Gram-negative rods, that were filtered with 0.22-, 1.0- and 5.0-micrometers filters, respectively. Rabbits in Groups 1 and 5 received uncontaminated and contaminated fluids, respectively, neither of which was filtered. Cultures were taken of blood and of fluid below the filters, and rabbit survival times were recorded. At 6, 8 and 12 hours, Group 2 survival time was not significantly different from that of the negative control; the survival times for Groups 3, 4 and 5 were significantly less than for the negative control. Fluid and blood cultures of Group 2 were negative or statistically indistinguishable from those of Group 1. Cultures for Groups 3, 4 and 5 were positive. The study suggests that 1.0- and 5.0-micrometers inline final filters have no beneficial effect on survival following infusion of contaminated fluids, but that 0.22-micrometer inline final filters increase survival time.     

Bacterial endotoxin retention by inline intravenous filters

Filters used in i.v. administration sets were tested for their ability to retain bacterial endotoxins for up to 96 hours of continuous infusion. Inline filters composed of cellulose ester, polyacrylate, polypropylene, polyethylene, or Posidyne Nylon 66 were used during continuous infusion of 5% dextrose injection at 83 mL/hr. One milliliter of inoculum containing 10(8) Escherichia coli was injected through a port upstream from the filter. A bacterial filter was used to monitor the sterility of effluent from the inline filters. The effluent was tested with limulus amebocyte lysate (LAL) that could detect endotoxin concentrations greater than 50 pg/mL. A control solution was monitored for viability of the bacteria throughout the course of the study, and positive endotoxin controls were used to confirm the sensitivity of the LAL. Samples of effluent were tested at 0, 4, 19, 24, 48, 72, and 96 hours. Effluent from all filters was sterile throughout the study. LAL assay indicated that only the effluent from filters containing Posidyne Nylon 66 was free of endotoxins for 96 hours. Effluent from the other filters contained endotoxins immediately after injection of the E. coli. Of the inline filters tested, only the one composed of Posidyne Nylon 66 was able to retain E. coli endotoxin for 96 hours. Further study is needed with E. coli and other microorganisms that are likely contaminants of i.v. infusions.     

Microbial retention characteristics of sterilizing-grade membrane filters with alginate substituted for oil-based products

For oil-based products, FDA recommends substitution of the oil with a compound which has similar viscosity and physical characteristics. In this study, a substitute for oil-based products was screened by measuring the viscosity and filterability, and examined for the presence of cell clumps in the various test fluids using an optical microscopy. The viscosity of the test fluids measured in the range of about 60-75 cP. Brevundimonas diminuta (formerly Pseudomonas diminuta), a standard challenge test organism for validation of 0.2 micron rated membrane filters, formed clumps in oils (corn, olive, sesame, and soybean) and polyethylene glycol (PEG, Molecular Weight (MW) = 400 and 1,000). During the viability test, cells suspended in 80% glycerol showed a ten-fold mortality rate after an exposure for 6 hours, but there was no significant change in viability in alginate (low, medium, and high viscosity) for 24 hours. These results suggested that alginate is better suited as a substitute for oil-based products than 80% glycerol. Since high viscosity fluids take longer to filter, the glycerol mortality rate would influence the challenge test negatively. A scaled-down filtration system has been developed for the described trials, and the bacterial challenge and bubble point tests have been performed in 1.6% alginate (66.7 cP), which was the choice of carrier fluid.     

Cytotoxicity assessment of chlorinated bacteria in water using the RNA synthesis inhibition method

Chlorination of drinking water containing organic materials is known to generate toxic by-products. We suggested that such compounds may also be produced by interactions between chlorine and bacteria present in water. To confirm this hypothesis, a method based on RNA synthesis inhibition of HeLa S3 human cells in the presence of toxic compounds was applied. This method is rapid and highly sensitive since the concentration of the samples is not required. Furthermore, it was shown to be a suitable method for measurement of the cytotoxicity of water. Aeromonas hydrophila suspensions, prepared with pyrodistilled water, devoid of any organic material, were chlorinated for a definite contact time. HeLa S3 cells were incubated (20 h, 37 degrees C) in a culture medium prepared with the chlorinated bacteria suspensions. The rate of incorporation of 3H uridine into RNA was used as a measure of RNA synthesis and was evaluated in the presence and absence of chlorinated bacteria suspension. This study showed that chlorinated bacteria suspensions are cytotoxic. We observed that 0.22 microm filters retain cytotoxic compounds but 0.45 microm filters did not. Chlorine concentration and bacteria level influence the cytotoxicity. First, the toxicity level increases with chlorine concentration, then it decreases when chlorine concentration is too high. On another hand, a dose effect relationship between bacteria concentration and cytotoxicity was established.     

Filter extrusion of liposomes using different devices: comparison of liposome size, encapsulation efficiency, and process characteristics.

Liposomes were prepared by stepwise extrusion through 5, 1, 0.4, 0.2, 0.1 and 0.05 microm pore sizes using two different filter-extruders, the continuous high pressure device Dispex Maximator (CE) or alternatively the discontinuous Avestin LiposoFast (DE). The liposome dispersions obtained were compared in terms of particle size, lamellarity and encapsulation efficiency of calcein. The liposomes were smaller with CE than DE at all stages due to higher flow rates and pressure drops, except for final filter pore size (0.05 microm) where both preparations had similar sizes. The particle size analysis technique itself had a strong influence on the liposome sizes measured. For bigger liposomes (extruded through 0.4 microm filters) the Nicomp 370 revealed bigger volume-based mean particle sizes along with more stringent differences between volume-based and number-based diameters than the Malvern Zetasizer. In contrast, for small liposomes extruded through 0.05 microm filters, similar liposome sizes were found no matter which of the two PCS techniques or cryo-transmission electron microscopy was used. In congruence to the liposome sizes measured, encapsulation efficiencies were smaller for CE than DE at all filter stages except the final (0.05 microm). No lipid loss occurred and lyso-phosphatidylcholine formation was negligible irrespective of which extrusion technique was used.