A Strategy for the Control of Catheter Blockage by Crystalline Proteus mirabilis Biofilm Using the Antibacterial AgentTriclosan

European Urology European Urology 48 (2005) 838 845 A Strategy for the Control of Catheter Blockage by Crystalline Proteus mirabilis Biofilm Using the Antibacterial AgentTriclosan G.Ll. Jones a,1, A.D. Russell b, Z. Caliskan a, D.J. Stickler a, * a Cardiff School of Biosciences, Cardiff University, Main Building, Park Place PO Box 915, Cardiff CF10 3TL, Wales, UK b Welsh School of Pharmacy, Cardiff University, Cardiff, UK Accepted 7 July 2005 Available online 3 August 2005 Abstract Objectives: Catheter blockage by crystalline Proteus mirabilis biofilm is a common complication in patients undergoing long-term indwelling bladder catheterisation. Previously we have shown that inflating the retention balloons of all-silicone catheters with triclosan solutions prevents the encrustation process. The aim of the present work was to examine whether this strategy is effective in latex-based catheters. Methods: Laboratory bladder models were fitted with catheters and the retention balloons inflated with water or various concentrations of triclosan. The urine was inoculated with Pr. mirabilis and the times catheters took to block recorded. Results: Control catheters blocked in mean times ranging from 18 to 27 h. The ph of the urine rose from 6.1 to >8.6. In models with latex-based catheters inflated with 1 10 mg/ml triclosan, the urinary ph was controlled, the numbers of organisms in the urine was reduced and the catheters drained freely for the 7 day experimental period. Electron microscopy confirmed that crystalline biofilm was blocking control catheters. Little sign of encrustation was visible on the test catheters. Conclusion: Inflating the retention balloons with triclosan could have practical applications in controlling encrustation on both latex and silicone-based catheters. # 2005 Elsevier B.V. All rights reserved. Keywords: Catheterisation; Proteus mirabilis; Biofilms; Urinary tract infections; Triclosan 1. Introduction Encrustation of catheters is a common complication in the care of many patients undergoing long-term indwelling bladder catheterisation [1,2]. The problem stems from infection by urease producing bacteria, particularly Proteus mirabilis [3 5]. These organisms colonize catheters producing biofilm communities embedded in a gel-like polysaccharide matrix. The * Corresponding author. Tel. +44 2920 874311/87 40 00; Fax: +44 2920 874305. E-mail address: stickler@cardiff.ac.uk (D.J. Stickler). 1 Gwennan Jones was part-funded by a postgraduate studentship from Cardiff University. There was no other funding source for this work. urease they produce hydrolyses urea to ammonia and carbon dioxide, elevating the ph of the urine and biofilm. Crystals of struvite and apatite form under the resulting alkaline conditions, and crystalline biofilm results. This material can occlude the catheter eyehole or drainage lumen and prevent the flow of urine from the bladder, seriously compromising patients health and welfare [6]. All types of Foley catheter are vulnerable to this problem and there are no effective procedures available for its control [7]. Conceptually, the simplest way to prevent the biofilm formation is to impregnate catheters with a broadspectrum antimicrobial agent that elutes into the surrounding environment. In this way planktonic 0302-2838/$ see front matter # 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.eururo.2005.07.004

G.L. Jones et al. / European Urology 48 (2005) 838 845 839 bacteria in the vicinity of the device could be attacked before they colonized the surface and adopt the biofilm resistant phenotype [8]. However, difficulties in delivering effective concentrations of antimicrobial agents from catheters for prolonged periods has limited the usefulness of antimicrobial catheters in patients undergoing long-term bladder management. Bibby and colleagues [9] suggested that the catheter retention balloon could be used as a reservoir for the delivery of antimicrobial agents into the catheterised bladder. Using an in-vitro model they showed that mandelic acid could diffuse through the balloon. Exploitation of the retention balloon in this way allows the catheter to be loaded with large quantities of an agent. The membrane of the balloon might also provide a diffusion barrier to control the release of active agents over protracted periods into the bladder urine. Inview of its importance in the process of encrustation and blockage of urinary catheters, we decided to develop a strategy that specifically targeted Pr. mirabilis. Although mandelic acid has previously been used to perform bladder washouts, its activity against Pr. mirabilis is poor [10]. In contrast,pr. mirabilis has been shown to be very sensitive to the biocide triclosan. The minimum inhibitory concentration of this agent for strains of Pr. mirabilis isolated from encrusted catheters has beenreportedas 0.5 mg/ml [11]. Using a laboratory model of the catheterised bladder supplied with an artificial urine and infected with Pr. mirabilis, we have previously shown that triclosan can diffuse through the balloons of all-silicone catheters and inhibit crystalline biofilm formation. In these tests, control catheters inflated with water blocked within 24 h, whereas the test catheters with balloons inflated with triclosan (10 mg/ml in 5% w/v polyethylene glycol) drained freely for at least seven days[12]. We now wishto reportthatthisstrategy is also effective with latex-based catheters and in experimental models supplied with pooled human urine. 2. Materials and methods 2.1. The bladder model The bladder model has been described previously [13]. It consists of a glass chamber maintained at 37 8C by a water jacket. Each model was sterilized by autoclaving and then a 14 Ch allsilicone catheter or a hydrogel-coated or silicone elastomer-coated latex catheter (Bard Ltd. Crawley, United Kingdom) inserted into the chamber through an outlet at the base. The catheter retention balloons were inflated with 10 ml of water or various concentrations of triclosan (CIBA, Basle, Switzerland). The catheters were connected to drainage tubes and bags in the normal way. Sterile urine was pumped into the chambers at 0.5 ml/min, so that residual volumes collected below the catheter eye-holes before flowing through the drainage tube to the collecting bags. 2.2. Preparation of the triclosan Triclosan is relatively insoluble in water. The formulations used in this study contained 0.01 to 10 mg/ml triclosan in 5% w/v polyethylene glycol (PEG) (Sigma, Poole, UK). This mixture was stirred overnight with heating to 70 8C to produce a stable white colloidal suspension. 2.3. The urine growth medium The artificial urine used in the experimental work was based on that devised by Griffith and colleagues [14]. Its composition and method of sterilization has been described previously [13]. In some experiments human urine was used. This was collected from healthy volunteers, pooled and sterilized by filtration. 2.4. Experimental protocol Sets of models were assembled and primed with urine up to the level of the eye-holes. The supply of urine was then switched off, 10 ml was removed from the bladder chamber and replaced by 10 ml of a 4 h culture of Pr. mirabilis B2, a strain that had been isolated from a patient s catheter and was capable of blocking catheters rapidly with crystalline bacterial biofilm. After an hour to allow the test organism to establish itself in the bladder urine, the urine supply was switched back on and the models operated for maximum periods of seven days or until the catheters blocked. The time taken for catheters to block was recorded, together with the ph of the urine at blockage. Viable bacterial cell counts of the residual urine at blockage were performed on CLED Agar (Oxoid, Ltd., Basingstoke, United Kingdom). 2.5. Scanning electron microscopy Catheters were removed from the models at the end of the experimental periods. Two sections (1 cm in length) were cut from each catheter, one included the eye-hole, the second from immediately below the eye-hole. These sections were viewed directly in a JEOL 5200 scanning electron microscope (Jeol, Ltd., Tokyo, Japan) using the low vacuum setting. 2.6. Chemical estimation of catheter encrustation Catheters were removed from the models, cut into 2 cm sections which were then soaked in nitric acid (4% v/v). These samples were then sonicated at 35 khz for 5 min in a sonic cleaning bath to facilitate the dissolution of the crystals embedded in the catheter biofilm. The resulting solution was assayed for calcium and magnesium using atomic absorption spectrophotometry. 2.7. Statistical analysis One-way ANOVA carried out at 95% confidence interval was the statistical test of choice. It was carried out using Minitab 1 release 13 software (Minitab, Inc., Pennsylvania, USA). If the assumptions required to perform ANOVA were violated, the Kruskal Wallis test was performed at 95% confidence interval. Where appropriate the standard error (SE) of the mean is indicated. 3. Results 3.1. The effect of triclosan on catheter blockage Bladder models fitted with all-silicone, hydrogelcoated latex or silicone-coated latex catheters were supplied with artificial urine. Test catheters were inflated with 0.01, 0.1, 0.5, 1 and 10 mg/ml triclosan

840 G.L. Jones et al. / European Urology 48 (2005) 838 845 Table 1 The effect of inflating the retention balloons with triclosan on the times catheters take to block Triclosan concentration (mg/ml) used to inflate balloons Mean SE times to catheter blockage (h) All-silicone (n) Hydrogel-coated latex (n) Silicone-coated latex (n) 0.0 (Water control) 27.1 3.0 (9) 22.9 2.6 (7) 18.8 1.9 (8) 0.0 (5% PEG control) 18.8 6.1 (3) 19.3 1.5 (3) 18.0 4.4 (3) 0.01 43.3 10.9 (3) 20.0 1.2 (3) 24.3 5.2 (3) 0.1 77.3 2.0 (3) ** 59.0 4.0 (3) ** 66.0 2.0 (3) ** 0.5 >168 (3) 133.0 6.0 (3) ** 141.7 7.5 (3) ** 1.0 >168 (3) >168 (3) >168 (3) * 10 >168 (6) >168 (3) >168 (3) Models were inoculated with Pr. mirabilis B2 and supplied with urine for up to seven days. The time at which each catheter blocked was noted. The number of replicates (n) for each experiment is indicated. * Two replicates drained freely for seven days, but the third blocked at 144 h. ** Indicates that the times to blockage were significantly longer than for the control catheters inflated with water (p < 0.001). in 5% PEG, control catheters were inflated with water or 5% PEG. The models were inoculated with the test strain of Pr. mirabilis and operated for a maximum period of seven days. The times the various catheters took to block were recorded and the results of replicated experiments are summarized in Table 1. The control catheters all blocked rapidly. For each type of catheter there was no significant difference between the times to blockage when water or PEG was used to inflate the retention balloons. When solutions containing triclosan at 0.5 to 10.0 mg/ml were used to inflate the balloons, the all-silicone catheters drained freely for the full experimental period. Slightly higher concentrations of triclosan were required to maintain normal drainage of the latex-based catheters. The ph of the urine in the control models increased from 6.1 to mean values above 8.6 at blockage (Fig. 1). Inflating the balloons with triclosan controlled this ph rise, at 10 mg/ml for example, in all cases the urinary ph remained acid for the duration of the test. The results presented in Fig. 2 show that inflating the retention balloons with triclosan also reduced the numbers of viable cells in the bladder urine. At 10 mg/ml, 4 log reductions in the viable cell counts of Pr. mirabilis were recorded for all catheter types. Scanning electron microscopy revealed that the eyeholes and the central channels of all the control catheters were blocked with crystalline material. In contrast very little encrustation was visible on catheters inflated with 10 mg/ml triclosan (Fig. 3). 3.2. Do catheters become impregnated with triclosan? A second set of experiments was performed to determine if inflating the catheter balloon with a triclosan solution resulted in the antibacterial agent Fig. 1. The effect of triclosan on the ph of the urine. The ph of the residual urine in the bladder model was measured at blockage or at seven days in those cases where the catheter drained freely for the full experimental period. The ph of the urine supplied to the model was 6.1. ( ) Indicates allsilicone catheters; ( ) hydrogel-coated latex catheters; ( ) silicone-coated latex catheters. Fig. 2. The effect of triclosan on the viable cell numbers of Pr. mirabilis in urine. The viable cell count (log 10 cfu/ml + 1) of bacteria in the residual urine was performed at catheter blockage or at seven days in those models where the catheter drained freely for the full experimental period. ( ) Indicates all-silicone catheters; ( ) hydrogel-coated latex catheters and ( ) silicone-coated latex catheters.

G.L. Jones et al. / European Urology 48 (2005) 838 845 841 Fig. 3. Scanning electron-micrographs of a control hydrogel-coated latex catheter inflated with water and a test catheter inflated with 10 mg/ml triclosan in 5% PEG that had been removed from Pr. mirabilis infected bladder models. (A) is of the eye-hole and (B) shows a cross-section taken just below the eye-hole of a control catheter that blocked at 30 h. (C and D) show the same views of a test catheter removed from the model after seven days. becoming impregnated in the catheter material. Catheters were inflated with 10 mg/ml triclosan in 5% PEG, inserted into sterile models and supplied with sterile artificial urine for 48 h. The catheters were deflated and washed thoroughly by inflating and deflating the balloon ten times with water. Sections (1 cm in length) were cut from the catheters and placed on TSA plates that had been spread with lawns of Pr. mirabilis B2. Following overnight incubation at 37 8C, the plates were observed for any zones of inhibition of growth of the test bacterium around the catheter sections. None of the sections of the silicone-coated latex catheter produced a zone of clearing on the bacterial lawn. Only the section from directly beneath the catheter balloons of the hydrogel-coated catheter produced a zone of inhibition. In the case of the all-silicone catheter however, all the sections tested produced zones of inhibition. These results are illustrated in Fig. 4. 3.3. The effect of triclosan on catheter encrustation in models supplied with human urine Bladder models fitted with all-silicone catheters were set up with test catheters inflated with 10 mg/ ml triclosan in 5% PEG and control catheters inflated with water. The models were inoculated with Pr. mirabilis B2 and supplied with sterile pooled human urine for 48 h. The amounts of calcium and magnesium deposited on each catheter were determined and the mean values calculated from three replicated experiments are shown in Fig. 5C. The extent of encrustation on a set of catheters from a fourth experiment was visualized by scanning electron microscopy. The mean ph of the pooled human urine at the beginning of the experiments was 6.6 0.1. None of the models blocked within the experimental period. The triclosan prevented the rise in urinary ph (Fig. 5A) and viable cell counts from the residual urine

842 G.L. Jones et al. / European Urology 48 (2005) 838 845 4. Discussion Fig. 4. The three types of catheter were inserted into bladder models, inflated with triclosan and the models supplied with artificial urine. After 48 h catheters were removed and washed thoroughly. Sections cut from these catheters were then placed on the surface of TSA plates that had been seeded with a lawn of the test strain of Pr. mirabilis B2. On incubation a large zone of inhibition of bacterial growth is visible around the section of all-silicone catheter but not around the latex-based catheters. of the test models at 48 h was significantly lower (p < 0.05) than from the control models (Fig. 5B). Analysis revealed that substantial amounts of calcium and magnesium had been deposited on the control catheters. Significantly (p < 0.001) less encrustation was recovered from the triclosan-treated catheters (Fig. 5C). Electron microscopy (Fig. 6) confirmed the heavier encrustation on the control catheters. An effective technique to control the blockage of catheters by crystalline Pr. mirabilis biofilm would be a major improvement in the care of patients undergoing long-term bladder catheterisation. Our previous study [12] using the Pr. mirabilis infected laboratory models of the catheterised bladder, showed that inflation of the retention balloon of all-silicone catheters with triclosan (10 mg/ml) prolonged the operational life of the catheters at least seven fold. The increase in urinary ph to >8.5 that accompanies Pr. mirabilis infection was inhibited and seven days after inoculation, when the experiments were terminated, the ph of the urine was still 6.2. The catheters were draining freely and showed no signs of the encrustation that blocked control catheters within 25 h. As many patients use hydrogel or silicone coated latex catheters rather than all-silicone devices, it is important to establish whether the triclosan strategy is also effective with these latex-based devices. The results presented in Table 1 confirm the activity of triclosan in the case of the all-silicone catheters. They show that triclosan at concentrations as low as 0.1 mg/ml significantly extend the life-time of the catheter (p < 0.001). At concentrations of 0.5 to 10 mg/ml the flow of urine through these catheters was maintained for the full seven days of the experiment. In the cases of the latex-based catheters, triclosan at 0.1 mg/ml also significantly increased the time catheters took to block (p < 0.001). Concentrations of 1 to 10 mg/ml were required however, to maintain the flow of urine for the duration of the test. The important requirement for controlling encrustation is to keep the ph of the urine below that at which Fig. 5. The effect of triclosan on catheter encrustation in models supplied with human urine. Sets of models fitted with all-silicone catheters that had been inflated with water or triclosan (10 mg/ml in PEG) were supplied with human urine for 48 h. (A) presents the mean ph values of urine in the bladder chambers at 48 h. The mean viable cell counts (log 10 cfu/ml + 1) of bacteria in the residual urine in control and test models are presented in (B). The amounts of encrustation recovered from catheters removed from the models at 48 h, expressed as mg of calcium and magnesium per catheter are recorded in (C). The mean values were calculated from four replicated experiments for A and B and three replicates for C.

G.L. Jones et al. / European Urology 48 (2005) 838 845 843 Fig. 6. Scanning electron-micrographs of a control all-silicone catheter inflated with water and a test catheter inflated with 10 mg/ml triclosan in 5% PEG from Pr. mirabilis infected bladder models supplied with human urine after 48 h. (A) is of the eye-hole and (B) shows a cross-section taken just below the eye-hole of a control catheter. (C and D) show the same views of a test catheter. calcium and magnesium salts come out of solution. This nucleation ph (phn), has been reported to be around 7.6 in catheterised patients [15]. Thus if the ph of urine can be maintained below 7.0 it is unlikely that crystalline biofilm will form. The results presented in Fig. 1 suggest that while concentrations of triclosan in the range of 1 to 10 mg/ml will keep the urinary ph in the safe zone for all-silicone and hydrogel-coated latex catheters, concentrations above 1 mg/ml are required to have this effect in the silicone-coated latex catheters. These effects and the observations that triclosan produced substantial reductions in the numbers of viable cells in the urine (Fig. 2) show that this biocide is capable of diffusing through the membranes of the retention balloons of both the silicone and latex-based catheters. The only major difference between the observations with the silicone and the latex-based catheters is that the triclosan does not impregnate the latex devices (Fig. 4). The conditions used in these experiments are intended to simulate conditions in which a catheter is introduced into a bladder that is heavily infected with a pure culture of Pr. mirabilis (10 8 cfu/ml) in urine at ph > 8.5. The concentrated urine and the low flow rate to the bladder also simulates conditions in an elderly patient who has a low fluid intake. Under these experimental conditions, catheter encrustation is rapid and catheters block more quickly than they do in most patients. The observations that after seven days the urine was still acidic and there was little sign of encrustation on the catheters, under severe test conditions that normally produce complete blockage of the catheters in around 24 h, suggest that the triclosan strategy should significantly extend the

844 G.L. Jones et al. / European Urology 48 (2005) 838 845 life-span of catheters in patients prone to encrustation. Having established the efficacy of the strategy in models supplied with a standard artificial urine, it was important to check that the effect is also manifest in the presence of human urine. The experiments reported in Fig. 5 were conducted over 48 h periods because of technical difficulties in maintaining a supply of the large volumes of sterile human urine required when the models are operated for seven days. The results (Fig. 5) show that the mean amounts of calcium and magnesium accumulating on the control catheters in 48 h was over 13 mg/catheter, compared to <0.3 mg/catheter when the balloons were inflated with triclosan. The reduction in the rate of crystalline biofilm formation to less than 3% of the control, in experiments performed in human urine, further supports the contention that this strategy should be effective in clinical practice. Although the control catheters did not block in the 48 h period, the scanning electron micrographs (Fig. 6) illustrate the extensive encrustation of their eye-holes and central channels. In contrast, the test catheter was clear of crystalline biofilm at 48 h. The possible effects of triclosan on the bladder epithelium, have not been tested, however this antibacterial agent is used in soaps, handwashes, deodorants, toothpastes and mouthwashes. Its safety has been established through extensive acute and long-term toxicity, carcinogenicity, reproduction and teratology studies and the USA Food and Drug Administration has approved its use in oral care products [16]. Triclosan has been used extensively in many antibacterial preparations for over 30 years and there has been little sign that the clinical or domestic use of this agent has led to the generation of resistant organisms [17 19]. The possible development of bacterial resistance however, is an issue that should always be considered when introducing a chemical agent to deal with an infection-associated problem. In any clinical trial or subsequent clinical use of this strategy to control catheter encrustation, it will be important to monitor the urinary flora for signs of the emergence of resistance to triclosan. It is important to state that not all the species responsible for catheter-associated urinary tract infections are sensitive to this biocide [20]. We are not suggesting that this strategy could be used to control catheter-associated urinary tract infections. All the 118 isolates of Pr. mirabilis we have tested however, have proved to be highly sensitive to this agent (MICs ranging from 0.1 to 0.3 mg/ml). The evidence we present indicates that triclosan is effective in inhibiting the formation of crystalline biofilms by Pr. mirabilis, the species most commonly responsible for catheter encrustation. These observations confirm and extend our previous conclusions that inflating catheter balloons with triclosan solutions rather than water could have applications in controlling a common problem that complicates the care of many patients enduring long-term indwelling catheterisation. The method does not disturb the integrity of the closed drainage system and might possibly be appropriate for the delivery of other agents and drugs into the catheterised bladder. The strategy now needs to be tested in a clinical trial. 5. Conclusion These results suggest that inflating the retention balloons with triclosan rather than water could have practical applications in controlling encrustation by crystalline Pr. mirabilis biofilm on both latex and silicone-based catheters. References [1] Cools HJ, Van der Meer JW. Restriction of long-term indwelling urethral catheterisation in the elderly. Br J Urol 1986;58:683 8. [2] Kohler-Ockmore J, Feneley RC. Long-term catheterization of the bladder: prevalence and morbidity. Br J Urol 1996;77:347 51. [3] Mobley HL, Warren JW. Urease-positive bacteriuria and obstruction of long-term urinary catheters. J Clin Microbiol 1987;25:2216 7. [4] Kunin CM. Blockage of urinary catheters: role of microorganisms and constituents of the urine on formation of encrustations. J Clin Epidemiol 1989;42:835 42. [5] Stickler D, Ganderton L, King J, Nettleton J, Winters C. Proteus mirabilis biofilms and the encrustation of urethral catheters. Urol Res 1993;21:407 11. [6] Kunin CM. Urinary tract infections: detection, prevention and management. 5th ed. Baltimore: Williams and Wilkins; 1997, pp 226-78. [7] Morris NS, Stickler DJ, Winters C. Which indwelling urethral catheters resist encrustation by Proteus mirabilis biofilms?. Br J Urol 1997;80:58 63. [8] Danese PN. Antibiofilm approaches: prevention of catheter colonization. Chem Biol 2002;9:873 80. [9] Bibby JM, Cox AJ, Hukins DW. Feasibility of preventing encrustation of urinary catheters. Cells Mater 1995;2:183 95. [10] King JB, 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 7. [11] Stickler DJ. Susceptibility of antibiotic-resistant Gram-negative bacteria to biocides: a perspective from the study of catheter biofilms. J Appl Microbiol 2002;92(Suppl):163 70. [12] Stickler DJ, Jones GL, Russell AD. Control of encrustation and blockage of Foley catheters. Lancet 2003;361:1435 7.

G.L. Jones et al. / European Urology 48 (2005) 838 845 845 [13] Stickler DJ, Morris NS, Winters C. Simple physical model to study formation and physiology of biofilms on urethral catheters. Methods Enzymol 1999;310:494 501. [14] Griffith DP, Musher DM, Itin C. Urease. The primary cause of infection-induced urinary stones. Invest Urol 1976;13:346 50. [15] Choong S, Wood S, Fry C, Whitfield H. Catheter associated urinary tract infection and encrustation. Int J Antimicrob Agents 2001;17:305 10. [16] Jones RD, Jampani HB, Newman JL, Lee AS. Triclosan: a review of its effectiveness and safety in healthcare settings. Am J Infect Control 2000;28:184 96. [17] Sreenivasan P, Gaffar A. Antiplaque biocides and bacterial resistance: a review. J Clin Periodontol 2002;29:965 74. [18] McBain AJ, Bartolo RG, Catrenich CE, Charbonneau D, Ledder RG, Price BB, et al. Exposure of sink drain microcosms to triclosan: population dynamics and antimicrobial susceptibility. Appl Environ Microbiol 2003;69:5433 42. [19] Russell AD. Whither triclosan? J Antimicrob Chemother 2004; 53:693 5. [20] Bharganava HN, Leonard PA. Triclosan: applications and safety. Am J Infect Control 1996;24:209 18.