- Email: [email protected]
Some Observations on the Diffusion of Antimicrobial Agents Through the Retention Balloons of Foley Catheters
Some Observations on the Diffusion of Antimicrobial Agents Through the Retention Balloons of Foley Catheters G. J. Williams and D. J. Stickler* From the Cardiff School of Biosciences, Cardiff University, Cardiff, Wales, United Kingdom
Purpose: We examined the ability of antimicrobial agents to diffuse through the retention balloons of urinary catheters and inhibit their encrustation by Proteus mirabilis. Materials and Methods: An agar diffusion screening test was developed to detect agents capable of diffusing through catheter balloons and inhibiting the growth of P. mirabilis. The effect of inflating the balloons with antibacterials on the ability of P. mirabilis to encrust catheters was tested in laboratory models of the catheterized bladder. Results: Of 18 antimicrobial agents active on P. mirabilis only mandelic acid, phenoxyethanol, nalidixic acid and triclosan diffused through all-silicone catheter balloons to produce zones of inhibition against P. mirabilis. Polyurethane balloons were permeable to gentamicin and fluoroquinolones. Experiments with silicone catheters showed that inflating balloons with mandelic acid (100 gm/l) or ciprofloxacin (10 gm/l) failed to extended the time at which catheters became blocked in models inoculated with P. mirabilis. However, nalidixic acid (50 gm/l) significantly extended the lifespan of catheters (p ⬍0.05). Triclosan (10 gm/l) prevented the increase in urinary pH that induces crystal formation and inhibited the formation of crystalline biofilm, enabling the catheters to drain freely for the full 7-day experimental period. Conclusions: Inflation of silicone catheter retention balloons with solutions of nalidixic acid or triclosan rather than water should be considered as strategies to control catheter encrustation. Polyurethane balloons are more permeable than silicone balloons to gentamicin and the fluoroquinolones, and they should be investigated as an alternative to silicone or latex in catheter manufacture. Key Words: bladder, urinary catheterization, Proteus mirabilis, anti-infective agents, biofilms
the lifetime of the device. Difficulties in delivering effective concentrations of antimicrobial agents from catheters for prolonged periods have limited the usefulness of antimicrobial catheters in patients undergoing long-term bladder management. Bibby et al suggested that the catheter retention balloon could be used as a large reservoir for the delivery of antimicrobial agents into the catheterized bladder.7 The idea was that the balloon membrane might also provide a diffusion barrier to control the release of active agents into residual bladder urine for protracted periods. Subsequent experiments in a catheterized bladder laboratory model showed that the biocide triclosan, which is extremely active on P. mirabilis, can diffuse through the balloons of allsilicone catheters and inhibit crystalline biofilm formation. In these tests and under conditions in which control catheters inflated with water became blocked within 24 hours the test catheters with balloons inflated with triclosan (10 mg/ml in 5% weight per volume PEG) still drained freely when the experiment was terminated after 7 days.8 The direct delivery of antimicrobial agents through the catheter balloon to the bladder has several potential advantages. It does not disrupt the integrity of the closed drainage system, it avoids the selection of resistant gut flora that occurs when drugs are received systemically and it does not require the manufacture of novel catheters. We identified other antibacterial agents that diffuse through retention balloons and inhibit the formation of P. mirabilis biofilms on catheters.
he care of many patients undergoing long-term indwelling bladder catheterization is complicated by the encrustation and blockage of the catheters.1 The problem results from infection by urease producing bacteria, particularly Proteus mirabilis.2,3 These organisms colonize catheters and produce enormous populations of cells, termed bacterial biofilms, embedded in a gel-like polysaccharide matrix. The bacteria hydrolyze urea to ammonia and carbon dioxide, increasing the pH of the urine and biofilm, generating conditions under which crystals of struvite and apatite precipitate from solution. The deposition of this material in the biofilm can occlude the catheter eyehole or lumen and prevent the flow of urine from the bladder, seriously compromising patient health and welfare.4,5 All types of Foley catheters are vulnerable to this problem and there are no effective procedures available for its control.5 A possible strategy to prevent encrustation by these crystalline bacterial biofilms is to impregnate catheters with an antimicrobial agent that elutes into the surrounding environment. In this way planktonic bacteria in the urine could be attacked before they colonized the catheter surface.6 The release of the agent must be sustained to prevent the bacterial activity that causes the increase in urinary pH during
T
Submitted for publication November 7, 2006. * Correspondence: School of Biosciences, Main College, Cardiff University, Cardiff CF10 3TL, Wales, United Kingdom (telephone: 1222 874311; FAX: 1222 874305; e-mail: [email protected]).
0022-5347/07/1782-0697/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
697
Vol. 178, 697-701, August 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.03.091
698
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS
MATERIALS AND METHODS Medium and Chemicals Culture medium was obtained from Oxoid Ltd., Basingstoke, United Kingdom. Artificial urine (pH 6.1) supplied to the bladder model was based on that described by Griffith et al.9 Its preparation was described previously.10 Triclosan (Irgasan DP300) was obtained from CIBA Specialty Chemicals, Basel, Switzerland. Other agents were obtained from Sigma Chemicals, Poole, United Kingdom. Catheters All-silicone catheters were obtained from Bard Ltd., Crawley, United Kingdom. Polyurethane balloons were provided by Microcuff GmbH, Weinheim, Germany. MIC Determinations Aliquots (10 l) of a 1:100 dilution of 4-hour tryptone soya broth cultures of test organisms (104 cfu) were dropped onto Iso-Sensitest agar plates (Oxoid) containing a range of antibacterial concentrations. MIC was determined as the lowest concentration of antibacterial that inhibited bacterial growth after overnight incubation at 37C. The P. mirabilis strains tested were obtained from encrusted patient catheters. Screening Test to Detect Diffusion of Agents Through Catheter Balloons The balloons were inflated with 10 ml water (controls) or solutions of antibacterial agents (tests). Test and control catheters were positioned in a 12 ⫻ 12-inch square plastic culture dish. Molten tryptone soya agar (500 ml) at 45C that had been seeded with P. mirabilis B2 was then poured into the dish. After overnight incubation at 37C the dishes were examined for zones of bacterial inhibition in the agar surrounding the catheters.
Low Vacuum Scanning Electron Microscopy Catheters removed from the models were examined under low vacuum using a 5200 LV Scanning Electron Microscope (Jeol Ltd., Tokyo, Japan) to assess the degree of encrustation. Sections 1 cm long were cut from each catheter, placed onto carbon discs on aluminum stubs and viewed without fixation. Statistical Analysis One-way ANOVA was used at the 95% significance level to test differences between the means of data sets. Minitab®, release 13 was used to perform the calculations and test the normality of distribution of residuals and homogeneity of variances of the data. When appropriate, the SEM is indicated. RESULTS Susceptibility of P. Mirabilis to Select Antimicrobial Agents and the Ability of Agents to Diffuse Through Catheter Balloons Table 1 lists the MICs of a range of antimicrobial agents against 8 P. mirabilis strains. Results confirmed the sensitivity of P. mirabilis to triclosan (MIC 0.1 to 0.2 mg/l). Data also showed that oxolinic acid (MIC 0.1 to 1 mg/l), gentamicin (MIC 0.4 to 1 mg/l), and the fluoroquinolones ciprofloxacin (MIC 0.1 mg/l), norfloxacin (MIC 0.1 to 1 mg/l) and ofloxacin (MIC 0.1 to 0.5 mg/l) were extremely active against P. mirabilis. Agar diffusion screening tests showed that of the 18 antimicrobial agents tested only mandelic acid, phenoxyethanol, nalidixic acid and triclosan produced zones of
TABLE 1. MIC values of antimicrobial agents against 8 P. mirabilis strains and their ability to diffuse through all-silicone catheter balloons Agar Diffusion Screen Test
The Bladder Model The catheterized bladder model was described previously.10 In essence it consists of a glass chamber maintained at 37C by a water jacket. Each model was sterilized by autoclaving and then a sterile size 14 all-silicone catheter was inserted aseptically into the chamber via an outlet in its base. Catheters were inflated with 10 ml water or antibacterial solution and then attached to drainage tubes and bags. Experimental Protocol Sets of models were assembled and artificial urine was pumped into the bladder chamber until it submerged the retention balloons. The urine supply was then halted and models were inoculated with 10 ml 4-hour artificial urine cultures of P. mirabilis B2 (108 cfu/ml), a strain capable of blocking catheters rapidly with crystalline bacterial biofilm. The models were left for 1 hour to enable the organisms to become established in the residual urine. Fresh artificial urine was then pumped into the chamber at 0.5 ml per minute and left to run for 168 hours or until the catheters became blocked. Time to catheter blockage was recorded. Urine samples were removed from the models at various times to measure pH and determine the viable bacterial cell counts.
Antimicrobial Agents
MIC (mg/l)
Ampicillin sodium salt 2–Greater than 1,000 Cephalexin hydrate 15 Nalidixic acid sodium 5–50 salt Pipemidic acid 3–5 Oxolinic acid 0.1–1 Norfloxacin* 0.1–1 Ofloxacin* 0.1–0.5 Neomycin sulphate 15 Gentamicin* 0.4–1 Rifampicin* 4–6 Tetracycline 20 hydrochloride Trimethoprim 1–400 Chloramphenicol 5–10 Thiamphenicol 25–30 Nitrofurazone 10–30 Chlorhexidine diacetate 15 Mandelic acid 1,000–3,000 Triclosan 0.1–0.2
Diffusion Concentration Through (gm/l) Balloon 10 10 10
No No Yes
5 1 5† 5† 100† 2 5 10
No No No No No No No No
100† 10 10† 1† 10 100 10†
No No No No No Yes Yes
* When polyurethane balloons were inflated with antibiotic and ciprofloxacin (MIC 0.1 mg/l, tested at 5 gm/l, 5% weight per volume PEG in the catheter balloon) and examined for activity by agar diffusion screen test, there was no evidence that rifampicin diffused through the balloon but clear growth inhibition zones encircled the balloons inflated with gentamicin, and the fluoroquinolone antibiotics ciprofloxacin, ofloxacin and norfloxacin. † PEG at 5%.
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS
699
FIG. 2. Effect of inflating retention balloons with antimicrobials on ability of P. mirabilis B2 to encrust and block all-silicone catheters. Mean time to blockage, calculated from 3 replicated experiments, is shown for water controls (open bars) and catheters inflated with antibacterial solutions (gray bars). For triclosan catheters did not block and drained freely at 168 hours. Statistical analysis was done to determine p values for mean test and control data.
FIG. 1. Results of some agar diffusion screening tests in all-silicone catheter balloons inflated with water (A), nalidixic acid (B) and triclosan (C), and polyurethane catheter inflated with norfloxacin (D). Inhibition zones can be seen in agar seeded with P. mirabilis B2.
growth inhibition around the all-silicone catheter balloons on lawns of P. mirabilis B2 (table 1 and fig. 1). In a second series of tests the ability of a limited number of agents to diffuse through polyurethane balloons was examined. Rifampicin was not capable of diffusing through these balloons but clear zones of growth inhibition encircled the balloons inflated with gentamicin, ciprofloxacin, ofloxacin and norfloxacin (fig. 1). Did Inflation of Retention Balloons With Antimicrobial Agents Inhibit Encrustation and Blockage of All-Silicone Catheters? Control models were fitted with all-silicone catheters inflated with sterile deionized water and test models were fitted with catheters inflated with antimicrobial solutions, including ciprofloxacin (10 gm/l in 10% PEG), mandelic acid (100 gm/l), nalidixic acid sodium salt (50 gm/l) or triclosan (10 gm/l in 5% PEG). The models were inoculated with P. mirabilis B2 and supplied with artificial urine for a maximum of 168 hours or until catheter blockage occurred. Figure 2 shows the mean time required by catheters to become blocked in 3 replicated experiments. Residual urine pH in the models at catheter blockage or after 168 hours for triclosan test models were measured. Table 2 lists mean results together with the mean number of viable cells recovered from these urine samples. Figure 2 shows that inflating all-silicone catheters with a ciprofloxacin or a mandelic acid solution did not significantly extend the time required by catheters to become blocked (p ⬎0.05). The results of mandelic acid are understandable since Bibby et al reported that, when solutions containing
100 gm/l were used to inflate catheter balloons, the concentration achieved in urine was only 0.1 gm/l, considerably lower than the MIC of 1 to 3 gm/l for recent catheter isolates of P. mirabilis (table 1).7 A significant increase in catheter lifespan occurred when catheters were inflated with nalidixic acid vs that in water controls (mean 58 vs 30 hours, p ⫽ 0.002). Inflating allsilicone catheter balloons with phenoxyethanol (98% volume per volume) also resulted in catheters draining urine for significantly longer than control catheters (data not shown). However, unfortunately this caused the balloons to distort and shrink, and undermined their function in retaining the catheter in the bladder. To inhibit encrustation it is important to keep the pH of urine below that at which calcium and magnesium salts come out of solution, that is nucleation pH, which is normally around 7.6 in catheterized patients.11 Table 2 shows that triclosan was the only agent to maintained acidic conditions in P. mirabilis urine cultures. This agent enabled the catheters to drain freely for the full 7 days of the test. Scanning electron microscopy confirmed that, when these catheters were removed from the model, they showed little or no sign of encrustation (fig. 3). Triclosan was also the only
TABLE 2. Effect of antimicrobials on urinary pH and log number of viable P. mirabilis cells in residual urine Antimicrobial pH:* Ciprofloxacin Mandelic acid Nalidixic acid Triclosan No. viable cells (cfu/ml): Ciprofloxacin Mandelic acid Nalidixic acid Triclosan
Mean ⫾ SE Control
Mean ⫾ SE Test
8.81 ⫾ 0.05 8.56 ⫾ 0.23 8.73 ⫾ 0.09 8.59 ⫾ 0.26
8.58 ⫾ 0.19 8.47 ⫾ 0.24 8.92 ⫾ 0.03 6.15 ⫾ 0.03
7.51 ⫾ 0.28 7.35 ⫾ 0.3 7.81 ⫾ 0.04 7.38 ⫾ 0.1
7.57 ⫾ 0.2 7.24 ⫾ 0.26 7.59 ⫾ 0.17 4.95†
Data from the results of 3 replicated experiments with triclosan model values obtained after 168 hours and all other values obtained at catheter blockage. * Models were supplied with urine at pH 6.1. † In 2 of 3 replicates no viable cells were recovered from urine at 168 hours (value indicated was recorded from 1 replicate only).
700
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS
FIG. 3. Scanning electron micrographs show catheters inflated with water or 10 mg/ml triclosan in 5% PEG. Note condition of eyelet and lumen of control (A and B), and test (C and D) catheters. Control catheter was removed from P. mirabilis B2 infected bladder model when it blocked at 40 hours. Test catheter still drained freely when it was removed at 168 hours.
agent to decrease the number of viable P. mirabilis cells in bladder urine (table 2). DISCUSSION Recent studies show that after urinary pH increases to the level at which crystals form P. mirabilis biofilm develops even on surfaces that normally inhibit bacterial cell attachment. Under alkaline conditions macroscopic aggregates of cells and crystals form in urine. These conglomerates gravitate to the catheter surface and initiate crystalline biofilm formation.12 If antibacterial agents are to be used in strategies to control catheter encrustation, it is essential that they should elute from the catheter into urine and inhibit the increase in pH that triggers encrustation. Antibacterials fixed in the catheter polymer are not likely to stop encrustation. For long-term catheters the problems are loading them with sufficient antimicrobial agent and then controlling its release. Bibby et al noted that mandelic acid could diffuse through the balloons of all-silicone catheters for at least 4 weeks.7 Although mandelic acid has been used in bladder instillations to clear bacteriuria in catheterized patients, its activity against P. mirabilis is poor.13 Triclosan has remarkable activity against isolates of this species from encrusted catheters (table 1).14 It is also capable of diffusing through catheter balloons and preventing crystalline P. mirabilis biofilm formation.8 Loading the balloon with 10 ml triclosan at 10 gm/l achieved a release rate of 115 g daily.15 Therefore, it should be possible to achieve effective concentrations in urine for the 84 days of the normal maximum catheter placement period. However, agar diffusion screening tests showed that of the 18 antimicrobial agents tested only mandelic acid, phenoxyethanol, nalidixic acid and triclosan were capable of diffusing through silicone catheter balloons (table 1 and fig. 1). Polyurethane retention balloons were freely permeable to gentamicin and the fluoroquinolone drugs. Another potential advantage of these balloons is that, since they are composed of thin films of polymer, they do not form the cuffs that can sometimes make it difficult to remove the silicone cath-
eters from the bladder. The devices used in the screening tests were balloons fitted with simple tubes. It would be worth considering manufacturing catheters with polyurethane balloons because they are clearly more permeable to antibacterials and would allow the delivery of various antibiotics and possibly other agents directly into the bladder. The ability of prototype catheters inflated with various agents to prevent encrustation could then be tested in bladder models. When considering the results presented (fig. 2), it must be emphasized that experiments in the laboratory model simulated conditions in which a catheter is introduced into a bladder that is heavily infected with a pure culture of P. mirabilis (108 cfu/ml) with urinary pH around 8.5. Under these conditions catheters become encrusted and block more rapidly than they normally do in patients.16 The fact that inflating catheters with nalidixic acid significantly increased the catheter lifespan under these worst case conditions suggests that it would be worth examining this strategy in patients. Triclosan was clearly the most effective agent. It maintained urinary pH at 6.1 and even at 7 days there was little sign of crystalline biofilm on the catheters (table 2 and fig. 3). P. mirabilis was chosen as the test organism for this study since there is substantial evidence that it is the major cause of catheter encrustation.2–5 Other urease producing organisms, such as Providencia and Morganella morganii, can be found in the urine of catheterized patients but they are less able to form crystalline biofilm on catheters.15 Triclosan was shown to be extremely active against Providencia stuartii but has little activity against M. morganii and does not inhibit biofilm formation on catheters by this organism.15 The safety of triclosan was established through extensive acute and long-term toxicity, carcinogenicity, reproduction and teratology studies, and the United States Food and Drug Administration approved its use in oral care products.17 It is commonly included in personal care and household products. Concerns have been expressed that its over exploitation would select for triclosan resistance and crossresistance to other antimicrobial agents.18 However, recent surveys and laboratory studies failed to establish a link between triclosan use and antibacterial resistance.19 Nonetheless, when introducing a biocide such as triclosan into a treatment regimen, it is important to monitor patient urinary flora for changes in triclosan and antibiotic susceptibilities. Recently Stickler et al described a simple sensor for P. mirabilis that could be located in the urine drainage bag and provide early warning that catheter encrustation was developing.20 Such a device could signal the need to invoke the triclosan strategy to prevent catheter blockage and the resulting clinical crises. Thus, the idea is not to use the triclosan strategy indiscriminately in all long-term cases, but rather to limit its exploitation to circumstances in which the early stages of catheter encrustation has been identified. In this way the risk of selecting resistant organisms would be decreased. CONCLUSIONS Inflating silicone catheter balloons with solutions of nalidixic acid or triclosan rather than water inhibits the forma-
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS tion of crystalline bacterial biofilms on the catheters. Polyurethane balloons are more permeable than silicone balloons to gentamicin and the fluoroquinolones, and they should be investigated as an alternative to silicone or latex in catheter manufacture. If the strategy of inflating Foley catheters with antibacterial agents proves be transferable from laboratory models to patients, it could significantly improve the care of the many patients who must currently endure the discomfort and risks associated with recurrent catheter blockage.
8. 9.
10.
11.
12.
Abbreviations and Acronyms MIC ⫽ minimum inhibitory concentration PEG ⫽ polyethylene glycol REFERENCES 1.
2.
3.
4.
5.
6. 7.
Kohler-Ockmore J and Feneley RCL: Long-term catheterization of the bladder: prevalence and morbidity. Br J Urol 1996; 77: 347. Kunin CM: Blockage of urinary catheters: role of microorganisms and constituents of the urine on formation of encrustations. J Clin Epidemiol 1989; 42: 835. Mobley HL and Warren JW: Urease-positive bacteriuria and obstruction of long-term urinary catheters. J Clin Microbiol 1987; 25: 2216. Kunin CM: Care of the catheter. In: Urinary Tract Infections: Detection, Prevention and Management, 5th ed. Baltimore: Williams & Wilkins Co 1997; chapt 9, pp 261–262. Morris NS, Stickler DJ and McLean RJ: The development of bacterial biofilms on indwelling urethral catheters. World J Urol 1999; 310: 345. Danese PN: Antibiofilm approaches: prevention of catheter colonization. Chem Biol 2002; 9: 873. Bibby JM, Cox AJ and Hukins DWL: Feasibility of preventing encrustation of urinary catheters. Cells Mater 1995; 2: 183.
13.
14.
15.
16.
17. 18. 19. 20.
701
Stickler DJ, Jones GL and Russell AD: Control of encrustation and blockage of Foley catheters. Lancet 2003; 361: 1435. Griffith DP, Musher DM and Itin C: Urease. The primary cause of infection-induced urinary stones. Invest Urol 1976; 13: 346. Stickler DJ, Morris NS and Winters C: Simple physical model to study formation and physiology of biofilms on urethral catheters. Methods Enzymol 1999; 310: 494. Choong SKS, Hallson P, Whitfield HN and Fry CH: The physicochemical basis of urinary catheter encrustation. BJU Int 1999; 83: 770. Stickler DJ, Lear JC, Morris NS, MacLeod SA, Downer A and Cadd DH: Observations on the adherence of Proteus mirabilis onto polymer surfaces. J Appl Microbiol 2006; 100: 1028. King JB and Stickler DJ: The effect of repeated instillations of antiseptics on catheter-associated urinary tract infections: a study in a physical model of the catheterized bladder. Urol Res 1992; 20: 403. Jones GL, Russell AD, Caliskan Z and Stickler DJ: A strategy for the control of catheter blockage by crystalline Proteus mirabilis biofilm using the antibacterial agent triclosan. Eur Urol 2005; 48: 838. Jones GL, Muller CT, O’Reilly M and Stickler DJ: Effect of triclosan on the development of bacterial biofilms by urinary tract pathogens on urinary catheters. J Antimicrob Chemother 2006; 57: 266. Mathur S, Suller MT, Stickler DJ and Feneley RCL: Prospective study of individuals with long-term urinary catheters colonized with Proteus species. BJU Int 2006; 97: 121. Bhargava HN and Leonard PA: Triclosan: applications and safety. Am J Infect Control 1996; 24: 209. Schweizer HP: Triclosan: a widely used biocide and its link to antibiotics. FEMS Microbiol Lett 2001; 202: 1. Russell AD: Whither triclosan? J Antimicrob Chemother 2004; 53: 693. Stickler DJ, Jones SM, Adusei GO and Waters MG: A sensor to detect the early stages in the development of crystalline Proteus mirabilis biofilm on indwelling bladder catheters. J Clin Microbiol 2006; 44: 1540.
Purpose: We examined the ability of antimicrobial agents to diffuse through the retention balloons of urinary catheters and inhibit their encrustation by Proteus mirabilis. Materials and Methods: An agar diffusion screening test was developed to detect agents capable of diffusing through catheter balloons and inhibiting the growth of P. mirabilis. The effect of inflating the balloons with antibacterials on the ability of P. mirabilis to encrust catheters was tested in laboratory models of the catheterized bladder. Results: Of 18 antimicrobial agents active on P. mirabilis only mandelic acid, phenoxyethanol, nalidixic acid and triclosan diffused through all-silicone catheter balloons to produce zones of inhibition against P. mirabilis. Polyurethane balloons were permeable to gentamicin and fluoroquinolones. Experiments with silicone catheters showed that inflating balloons with mandelic acid (100 gm/l) or ciprofloxacin (10 gm/l) failed to extended the time at which catheters became blocked in models inoculated with P. mirabilis. However, nalidixic acid (50 gm/l) significantly extended the lifespan of catheters (p ⬍0.05). Triclosan (10 gm/l) prevented the increase in urinary pH that induces crystal formation and inhibited the formation of crystalline biofilm, enabling the catheters to drain freely for the full 7-day experimental period. Conclusions: Inflation of silicone catheter retention balloons with solutions of nalidixic acid or triclosan rather than water should be considered as strategies to control catheter encrustation. Polyurethane balloons are more permeable than silicone balloons to gentamicin and the fluoroquinolones, and they should be investigated as an alternative to silicone or latex in catheter manufacture. Key Words: bladder, urinary catheterization, Proteus mirabilis, anti-infective agents, biofilms
the lifetime of the device. Difficulties in delivering effective concentrations of antimicrobial agents from catheters for prolonged periods have limited the usefulness of antimicrobial catheters in patients undergoing long-term bladder management. Bibby et al suggested that the catheter retention balloon could be used as a large reservoir for the delivery of antimicrobial agents into the catheterized bladder.7 The idea was that the balloon membrane might also provide a diffusion barrier to control the release of active agents into residual bladder urine for protracted periods. Subsequent experiments in a catheterized bladder laboratory model showed that the biocide triclosan, which is extremely active on P. mirabilis, can diffuse through the balloons of allsilicone catheters and inhibit crystalline biofilm formation. In these tests and under conditions in which control catheters inflated with water became blocked within 24 hours the test catheters with balloons inflated with triclosan (10 mg/ml in 5% weight per volume PEG) still drained freely when the experiment was terminated after 7 days.8 The direct delivery of antimicrobial agents through the catheter balloon to the bladder has several potential advantages. It does not disrupt the integrity of the closed drainage system, it avoids the selection of resistant gut flora that occurs when drugs are received systemically and it does not require the manufacture of novel catheters. We identified other antibacterial agents that diffuse through retention balloons and inhibit the formation of P. mirabilis biofilms on catheters.
he care of many patients undergoing long-term indwelling bladder catheterization is complicated by the encrustation and blockage of the catheters.1 The problem results from infection by urease producing bacteria, particularly Proteus mirabilis.2,3 These organisms colonize catheters and produce enormous populations of cells, termed bacterial biofilms, embedded in a gel-like polysaccharide matrix. The bacteria hydrolyze urea to ammonia and carbon dioxide, increasing the pH of the urine and biofilm, generating conditions under which crystals of struvite and apatite precipitate from solution. The deposition of this material in the biofilm can occlude the catheter eyehole or lumen and prevent the flow of urine from the bladder, seriously compromising patient health and welfare.4,5 All types of Foley catheters are vulnerable to this problem and there are no effective procedures available for its control.5 A possible strategy to prevent encrustation by these crystalline bacterial biofilms is to impregnate catheters with an antimicrobial agent that elutes into the surrounding environment. In this way planktonic bacteria in the urine could be attacked before they colonized the catheter surface.6 The release of the agent must be sustained to prevent the bacterial activity that causes the increase in urinary pH during
T
Submitted for publication November 7, 2006. * Correspondence: School of Biosciences, Main College, Cardiff University, Cardiff CF10 3TL, Wales, United Kingdom (telephone: 1222 874311; FAX: 1222 874305; e-mail: [email protected]).
0022-5347/07/1782-0697/0 THE JOURNAL OF UROLOGY® Copyright © 2007 by AMERICAN UROLOGICAL ASSOCIATION
697
Vol. 178, 697-701, August 2007 Printed in U.S.A. DOI:10.1016/j.juro.2007.03.091
698
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS
MATERIALS AND METHODS Medium and Chemicals Culture medium was obtained from Oxoid Ltd., Basingstoke, United Kingdom. Artificial urine (pH 6.1) supplied to the bladder model was based on that described by Griffith et al.9 Its preparation was described previously.10 Triclosan (Irgasan DP300) was obtained from CIBA Specialty Chemicals, Basel, Switzerland. Other agents were obtained from Sigma Chemicals, Poole, United Kingdom. Catheters All-silicone catheters were obtained from Bard Ltd., Crawley, United Kingdom. Polyurethane balloons were provided by Microcuff GmbH, Weinheim, Germany. MIC Determinations Aliquots (10 l) of a 1:100 dilution of 4-hour tryptone soya broth cultures of test organisms (104 cfu) were dropped onto Iso-Sensitest agar plates (Oxoid) containing a range of antibacterial concentrations. MIC was determined as the lowest concentration of antibacterial that inhibited bacterial growth after overnight incubation at 37C. The P. mirabilis strains tested were obtained from encrusted patient catheters. Screening Test to Detect Diffusion of Agents Through Catheter Balloons The balloons were inflated with 10 ml water (controls) or solutions of antibacterial agents (tests). Test and control catheters were positioned in a 12 ⫻ 12-inch square plastic culture dish. Molten tryptone soya agar (500 ml) at 45C that had been seeded with P. mirabilis B2 was then poured into the dish. After overnight incubation at 37C the dishes were examined for zones of bacterial inhibition in the agar surrounding the catheters.
Low Vacuum Scanning Electron Microscopy Catheters removed from the models were examined under low vacuum using a 5200 LV Scanning Electron Microscope (Jeol Ltd., Tokyo, Japan) to assess the degree of encrustation. Sections 1 cm long were cut from each catheter, placed onto carbon discs on aluminum stubs and viewed without fixation. Statistical Analysis One-way ANOVA was used at the 95% significance level to test differences between the means of data sets. Minitab®, release 13 was used to perform the calculations and test the normality of distribution of residuals and homogeneity of variances of the data. When appropriate, the SEM is indicated. RESULTS Susceptibility of P. Mirabilis to Select Antimicrobial Agents and the Ability of Agents to Diffuse Through Catheter Balloons Table 1 lists the MICs of a range of antimicrobial agents against 8 P. mirabilis strains. Results confirmed the sensitivity of P. mirabilis to triclosan (MIC 0.1 to 0.2 mg/l). Data also showed that oxolinic acid (MIC 0.1 to 1 mg/l), gentamicin (MIC 0.4 to 1 mg/l), and the fluoroquinolones ciprofloxacin (MIC 0.1 mg/l), norfloxacin (MIC 0.1 to 1 mg/l) and ofloxacin (MIC 0.1 to 0.5 mg/l) were extremely active against P. mirabilis. Agar diffusion screening tests showed that of the 18 antimicrobial agents tested only mandelic acid, phenoxyethanol, nalidixic acid and triclosan produced zones of
TABLE 1. MIC values of antimicrobial agents against 8 P. mirabilis strains and their ability to diffuse through all-silicone catheter balloons Agar Diffusion Screen Test
The Bladder Model The catheterized bladder model was described previously.10 In essence it consists of a glass chamber maintained at 37C by a water jacket. Each model was sterilized by autoclaving and then a sterile size 14 all-silicone catheter was inserted aseptically into the chamber via an outlet in its base. Catheters were inflated with 10 ml water or antibacterial solution and then attached to drainage tubes and bags. Experimental Protocol Sets of models were assembled and artificial urine was pumped into the bladder chamber until it submerged the retention balloons. The urine supply was then halted and models were inoculated with 10 ml 4-hour artificial urine cultures of P. mirabilis B2 (108 cfu/ml), a strain capable of blocking catheters rapidly with crystalline bacterial biofilm. The models were left for 1 hour to enable the organisms to become established in the residual urine. Fresh artificial urine was then pumped into the chamber at 0.5 ml per minute and left to run for 168 hours or until the catheters became blocked. Time to catheter blockage was recorded. Urine samples were removed from the models at various times to measure pH and determine the viable bacterial cell counts.
Antimicrobial Agents
MIC (mg/l)
Ampicillin sodium salt 2–Greater than 1,000 Cephalexin hydrate 15 Nalidixic acid sodium 5–50 salt Pipemidic acid 3–5 Oxolinic acid 0.1–1 Norfloxacin* 0.1–1 Ofloxacin* 0.1–0.5 Neomycin sulphate 15 Gentamicin* 0.4–1 Rifampicin* 4–6 Tetracycline 20 hydrochloride Trimethoprim 1–400 Chloramphenicol 5–10 Thiamphenicol 25–30 Nitrofurazone 10–30 Chlorhexidine diacetate 15 Mandelic acid 1,000–3,000 Triclosan 0.1–0.2
Diffusion Concentration Through (gm/l) Balloon 10 10 10
No No Yes
5 1 5† 5† 100† 2 5 10
No No No No No No No No
100† 10 10† 1† 10 100 10†
No No No No No Yes Yes
* When polyurethane balloons were inflated with antibiotic and ciprofloxacin (MIC 0.1 mg/l, tested at 5 gm/l, 5% weight per volume PEG in the catheter balloon) and examined for activity by agar diffusion screen test, there was no evidence that rifampicin diffused through the balloon but clear growth inhibition zones encircled the balloons inflated with gentamicin, and the fluoroquinolone antibiotics ciprofloxacin, ofloxacin and norfloxacin. † PEG at 5%.
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS
699
FIG. 2. Effect of inflating retention balloons with antimicrobials on ability of P. mirabilis B2 to encrust and block all-silicone catheters. Mean time to blockage, calculated from 3 replicated experiments, is shown for water controls (open bars) and catheters inflated with antibacterial solutions (gray bars). For triclosan catheters did not block and drained freely at 168 hours. Statistical analysis was done to determine p values for mean test and control data.
FIG. 1. Results of some agar diffusion screening tests in all-silicone catheter balloons inflated with water (A), nalidixic acid (B) and triclosan (C), and polyurethane catheter inflated with norfloxacin (D). Inhibition zones can be seen in agar seeded with P. mirabilis B2.
growth inhibition around the all-silicone catheter balloons on lawns of P. mirabilis B2 (table 1 and fig. 1). In a second series of tests the ability of a limited number of agents to diffuse through polyurethane balloons was examined. Rifampicin was not capable of diffusing through these balloons but clear zones of growth inhibition encircled the balloons inflated with gentamicin, ciprofloxacin, ofloxacin and norfloxacin (fig. 1). Did Inflation of Retention Balloons With Antimicrobial Agents Inhibit Encrustation and Blockage of All-Silicone Catheters? Control models were fitted with all-silicone catheters inflated with sterile deionized water and test models were fitted with catheters inflated with antimicrobial solutions, including ciprofloxacin (10 gm/l in 10% PEG), mandelic acid (100 gm/l), nalidixic acid sodium salt (50 gm/l) or triclosan (10 gm/l in 5% PEG). The models were inoculated with P. mirabilis B2 and supplied with artificial urine for a maximum of 168 hours or until catheter blockage occurred. Figure 2 shows the mean time required by catheters to become blocked in 3 replicated experiments. Residual urine pH in the models at catheter blockage or after 168 hours for triclosan test models were measured. Table 2 lists mean results together with the mean number of viable cells recovered from these urine samples. Figure 2 shows that inflating all-silicone catheters with a ciprofloxacin or a mandelic acid solution did not significantly extend the time required by catheters to become blocked (p ⬎0.05). The results of mandelic acid are understandable since Bibby et al reported that, when solutions containing
100 gm/l were used to inflate catheter balloons, the concentration achieved in urine was only 0.1 gm/l, considerably lower than the MIC of 1 to 3 gm/l for recent catheter isolates of P. mirabilis (table 1).7 A significant increase in catheter lifespan occurred when catheters were inflated with nalidixic acid vs that in water controls (mean 58 vs 30 hours, p ⫽ 0.002). Inflating allsilicone catheter balloons with phenoxyethanol (98% volume per volume) also resulted in catheters draining urine for significantly longer than control catheters (data not shown). However, unfortunately this caused the balloons to distort and shrink, and undermined their function in retaining the catheter in the bladder. To inhibit encrustation it is important to keep the pH of urine below that at which calcium and magnesium salts come out of solution, that is nucleation pH, which is normally around 7.6 in catheterized patients.11 Table 2 shows that triclosan was the only agent to maintained acidic conditions in P. mirabilis urine cultures. This agent enabled the catheters to drain freely for the full 7 days of the test. Scanning electron microscopy confirmed that, when these catheters were removed from the model, they showed little or no sign of encrustation (fig. 3). Triclosan was also the only
TABLE 2. Effect of antimicrobials on urinary pH and log number of viable P. mirabilis cells in residual urine Antimicrobial pH:* Ciprofloxacin Mandelic acid Nalidixic acid Triclosan No. viable cells (cfu/ml): Ciprofloxacin Mandelic acid Nalidixic acid Triclosan
Mean ⫾ SE Control
Mean ⫾ SE Test
8.81 ⫾ 0.05 8.56 ⫾ 0.23 8.73 ⫾ 0.09 8.59 ⫾ 0.26
8.58 ⫾ 0.19 8.47 ⫾ 0.24 8.92 ⫾ 0.03 6.15 ⫾ 0.03
7.51 ⫾ 0.28 7.35 ⫾ 0.3 7.81 ⫾ 0.04 7.38 ⫾ 0.1
7.57 ⫾ 0.2 7.24 ⫾ 0.26 7.59 ⫾ 0.17 4.95†
Data from the results of 3 replicated experiments with triclosan model values obtained after 168 hours and all other values obtained at catheter blockage. * Models were supplied with urine at pH 6.1. † In 2 of 3 replicates no viable cells were recovered from urine at 168 hours (value indicated was recorded from 1 replicate only).
700
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS
FIG. 3. Scanning electron micrographs show catheters inflated with water or 10 mg/ml triclosan in 5% PEG. Note condition of eyelet and lumen of control (A and B), and test (C and D) catheters. Control catheter was removed from P. mirabilis B2 infected bladder model when it blocked at 40 hours. Test catheter still drained freely when it was removed at 168 hours.
agent to decrease the number of viable P. mirabilis cells in bladder urine (table 2). DISCUSSION Recent studies show that after urinary pH increases to the level at which crystals form P. mirabilis biofilm develops even on surfaces that normally inhibit bacterial cell attachment. Under alkaline conditions macroscopic aggregates of cells and crystals form in urine. These conglomerates gravitate to the catheter surface and initiate crystalline biofilm formation.12 If antibacterial agents are to be used in strategies to control catheter encrustation, it is essential that they should elute from the catheter into urine and inhibit the increase in pH that triggers encrustation. Antibacterials fixed in the catheter polymer are not likely to stop encrustation. For long-term catheters the problems are loading them with sufficient antimicrobial agent and then controlling its release. Bibby et al noted that mandelic acid could diffuse through the balloons of all-silicone catheters for at least 4 weeks.7 Although mandelic acid has been used in bladder instillations to clear bacteriuria in catheterized patients, its activity against P. mirabilis is poor.13 Triclosan has remarkable activity against isolates of this species from encrusted catheters (table 1).14 It is also capable of diffusing through catheter balloons and preventing crystalline P. mirabilis biofilm formation.8 Loading the balloon with 10 ml triclosan at 10 gm/l achieved a release rate of 115 g daily.15 Therefore, it should be possible to achieve effective concentrations in urine for the 84 days of the normal maximum catheter placement period. However, agar diffusion screening tests showed that of the 18 antimicrobial agents tested only mandelic acid, phenoxyethanol, nalidixic acid and triclosan were capable of diffusing through silicone catheter balloons (table 1 and fig. 1). Polyurethane retention balloons were freely permeable to gentamicin and the fluoroquinolone drugs. Another potential advantage of these balloons is that, since they are composed of thin films of polymer, they do not form the cuffs that can sometimes make it difficult to remove the silicone cath-
eters from the bladder. The devices used in the screening tests were balloons fitted with simple tubes. It would be worth considering manufacturing catheters with polyurethane balloons because they are clearly more permeable to antibacterials and would allow the delivery of various antibiotics and possibly other agents directly into the bladder. The ability of prototype catheters inflated with various agents to prevent encrustation could then be tested in bladder models. When considering the results presented (fig. 2), it must be emphasized that experiments in the laboratory model simulated conditions in which a catheter is introduced into a bladder that is heavily infected with a pure culture of P. mirabilis (108 cfu/ml) with urinary pH around 8.5. Under these conditions catheters become encrusted and block more rapidly than they normally do in patients.16 The fact that inflating catheters with nalidixic acid significantly increased the catheter lifespan under these worst case conditions suggests that it would be worth examining this strategy in patients. Triclosan was clearly the most effective agent. It maintained urinary pH at 6.1 and even at 7 days there was little sign of crystalline biofilm on the catheters (table 2 and fig. 3). P. mirabilis was chosen as the test organism for this study since there is substantial evidence that it is the major cause of catheter encrustation.2–5 Other urease producing organisms, such as Providencia and Morganella morganii, can be found in the urine of catheterized patients but they are less able to form crystalline biofilm on catheters.15 Triclosan was shown to be extremely active against Providencia stuartii but has little activity against M. morganii and does not inhibit biofilm formation on catheters by this organism.15 The safety of triclosan was established through extensive acute and long-term toxicity, carcinogenicity, reproduction and teratology studies, and the United States Food and Drug Administration approved its use in oral care products.17 It is commonly included in personal care and household products. Concerns have been expressed that its over exploitation would select for triclosan resistance and crossresistance to other antimicrobial agents.18 However, recent surveys and laboratory studies failed to establish a link between triclosan use and antibacterial resistance.19 Nonetheless, when introducing a biocide such as triclosan into a treatment regimen, it is important to monitor patient urinary flora for changes in triclosan and antibiotic susceptibilities. Recently Stickler et al described a simple sensor for P. mirabilis that could be located in the urine drainage bag and provide early warning that catheter encrustation was developing.20 Such a device could signal the need to invoke the triclosan strategy to prevent catheter blockage and the resulting clinical crises. Thus, the idea is not to use the triclosan strategy indiscriminately in all long-term cases, but rather to limit its exploitation to circumstances in which the early stages of catheter encrustation has been identified. In this way the risk of selecting resistant organisms would be decreased. CONCLUSIONS Inflating silicone catheter balloons with solutions of nalidixic acid or triclosan rather than water inhibits the forma-
ANTIMICROBIAL DIFFUSION THROUGH CATHETER RETENTION BALLOONS tion of crystalline bacterial biofilms on the catheters. Polyurethane balloons are more permeable than silicone balloons to gentamicin and the fluoroquinolones, and they should be investigated as an alternative to silicone or latex in catheter manufacture. If the strategy of inflating Foley catheters with antibacterial agents proves be transferable from laboratory models to patients, it could significantly improve the care of the many patients who must currently endure the discomfort and risks associated with recurrent catheter blockage.
8. 9.
10.
11.
12.
Abbreviations and Acronyms MIC ⫽ minimum inhibitory concentration PEG ⫽ polyethylene glycol REFERENCES 1.
2.
3.
4.
5.
6. 7.
Kohler-Ockmore J and Feneley RCL: Long-term catheterization of the bladder: prevalence and morbidity. Br J Urol 1996; 77: 347. Kunin CM: Blockage of urinary catheters: role of microorganisms and constituents of the urine on formation of encrustations. J Clin Epidemiol 1989; 42: 835. Mobley HL and Warren JW: Urease-positive bacteriuria and obstruction of long-term urinary catheters. J Clin Microbiol 1987; 25: 2216. Kunin CM: Care of the catheter. In: Urinary Tract Infections: Detection, Prevention and Management, 5th ed. Baltimore: Williams & Wilkins Co 1997; chapt 9, pp 261–262. Morris NS, Stickler DJ and McLean RJ: The development of bacterial biofilms on indwelling urethral catheters. World J Urol 1999; 310: 345. Danese PN: Antibiofilm approaches: prevention of catheter colonization. Chem Biol 2002; 9: 873. Bibby JM, Cox AJ and Hukins DWL: Feasibility of preventing encrustation of urinary catheters. Cells Mater 1995; 2: 183.
13.
14.
15.
16.
17. 18. 19. 20.
701
Stickler DJ, Jones GL and Russell AD: Control of encrustation and blockage of Foley catheters. Lancet 2003; 361: 1435. Griffith DP, Musher DM and Itin C: Urease. The primary cause of infection-induced urinary stones. Invest Urol 1976; 13: 346. Stickler DJ, Morris NS and Winters C: Simple physical model to study formation and physiology of biofilms on urethral catheters. Methods Enzymol 1999; 310: 494. Choong SKS, Hallson P, Whitfield HN and Fry CH: The physicochemical basis of urinary catheter encrustation. BJU Int 1999; 83: 770. Stickler DJ, Lear JC, Morris NS, MacLeod SA, Downer A and Cadd DH: Observations on the adherence of Proteus mirabilis onto polymer surfaces. J Appl Microbiol 2006; 100: 1028. King JB and Stickler DJ: The effect of repeated instillations of antiseptics on catheter-associated urinary tract infections: a study in a physical model of the catheterized bladder. Urol Res 1992; 20: 403. Jones GL, Russell AD, Caliskan Z and Stickler DJ: A strategy for the control of catheter blockage by crystalline Proteus mirabilis biofilm using the antibacterial agent triclosan. Eur Urol 2005; 48: 838. Jones GL, Muller CT, O’Reilly M and Stickler DJ: Effect of triclosan on the development of bacterial biofilms by urinary tract pathogens on urinary catheters. J Antimicrob Chemother 2006; 57: 266. Mathur S, Suller MT, Stickler DJ and Feneley RCL: Prospective study of individuals with long-term urinary catheters colonized with Proteus species. BJU Int 2006; 97: 121. Bhargava HN and Leonard PA: Triclosan: applications and safety. Am J Infect Control 1996; 24: 209. Schweizer HP: Triclosan: a widely used biocide and its link to antibiotics. FEMS Microbiol Lett 2001; 202: 1. Russell AD: Whither triclosan? J Antimicrob Chemother 2004; 53: 693. Stickler DJ, Jones SM, Adusei GO and Waters MG: A sensor to detect the early stages in the development of crystalline Proteus mirabilis biofilm on indwelling bladder catheters. J Clin Microbiol 2006; 44: 1540.