861
The Use of Nonsteroidal Antiinflammatory Drugs to Prevent Adherence of Staphylococcus epidermidis to Medical Polymers Bruce F. Farber and Arlene G. Wolff
Division of Infectious Diseases. Department of Medicine. North Shore University Hospital-Cornell University Medical College, Manhasset. New York
The effect of salicylates and other nonsteroidal antiinflammatory drugs on the production of Staphylococcus epidermidis extracellular slime was studied. A dose-related decrease in slime production was observed with increasing concentrations of salicylic acid. S. epidermidis grown in 5 mM salicylic acid were less likely to adhere to Silastic, polyvinyl chloride, polyurethane, and Teflon catheters (P < .006); strains grown in 2 mM salicylic acid, ibuprofen, indomethacin, or phenylbutazone were less adherent to Silastic catheters (P < .001). Similar results were obtained with polyurethane catheters. S. epidermidis strains were less likely to adhere (43%-82% inhibition) to polyurethane catheters treated with 500 mM salicylic acid diluted in ethanol (P < .0001). Similar differences were not observed with acetaminophen, ibuprofen, or acetylsalicylic acid. Adherence of radiolabeled S. epidermidis to salicylic acid-treated Silastic catheters demonstrated a dose-related reduction. The use of salicylic acid to coat medical devices may decrease the incidence of device-related infection. Staphylococcus epidermidis has emerged as a major pathogen over the past decade. Previously considered contaminants, these organisms are the most common cause of prosthetic device- and catheter-related infection [1]. Certain strains of S. epidermidis produce an extracellular slime-like substance that appears to be an important virulence factor [2-4]. Recent studies have characterized the composition of S. epidermidis slime and its role in promoting the adherence of bacteria to polymers [5-7]. Extracellular slime produced by S. epidermidis is predominantly made of polysaccharides rich in glucose and galactose [6, 7]. Domenico et a1. [8] demonstrated that sodium salicylate reduces the amount of capsular polysaccharide produced by Klebsiella pneumoniae. Because S. epidermidis extracellular slime is a form ofexuberent capsular polysaccharide, we studied the effect of salicylates on slime production. We also studied the effect ofcoating medical polymers with nonsteroidal antiinflammatory drugs (NSAIDs) on bacterial adherence. These observations may have utility in the prevention of device-related infections.
Methods Preparation of extracts. Coagulase-negative staphylococci previously identified as S. epidermidis [9] were used in these
Received 18 February 1992; revised 4 May 1992. Financial support: Jane and Dayton Brown and Dayton Brown Jr. family fund. Reprints or correspondence: Dr. Bruce F. Farber, North Shore University Hospital-Cornell University Medical College, 300 Community Dr., Manhasset. NY I 1030.
The Journal of Infectious Diseases 1992;166:861-5 © 1992 by The University of Chicago. All rights reserved. 0022-1899/92/6604-0025$01.00
experiments. All strains had been found to be slime-positive by a qualitative method [10]. Crude slime filtrates were prepared from strains grown in chemically defined medium supplemented with various concentrations of sodium salicylate. The cultures were incubated for 48 h with shaking at 37°C and were centrifuged for 30 min at 27,000 g. The supernatant was dialyzed against sterile water for 48 h and then lyophilized. The freeze-dried product was reconstituted in 1/ I0 saturated sodium acetate, subjected to phenol and chloroform extractions as previously described, lyophilized, and weighed [9]. Spectrophotometric method. Adherence of coagulase-negative staphylococci to plastic culture plates was quantitatively determined by the method ofChristensen et a1. [10]. Organisms were grown in trypticase soy broth (TSB) for 18 h at 37°C. The cultures were then diluted I: I00 with fresh TSB with and without various concentrations of sodium salicylate. Aliquots (0.2 mL) of the diluted culture were pipetted into individual wells of sterile, polystyrene, 96-well flat-bottomed tissue culture plates (Corning Science Products, Medfield, MA). The plates were incubated for 18 h at 37°C. The contents of each well were gently aspirated and the well was washed four times with 0.2 mL of PBS. Adherent organisms were fixed with Bouin's fixative and then stained with Hucker crystal violet. Excess stain was rinsed off by placing the plate under running water. After drying. the optical densities of the stained adherent bacterial films were read (Micro ELISA Auto Reader, MR 600; Dynatech, Alexandria, VA) using a wavelength of 570 nm on the single-wavelength mode. Colony counts were also done from broth cultures in the presence ofincreasing concentrations ofNSAIDs. The cultures were grown in TSB. Adherence ofbacteria to catheter segments. TSB was prepared at 2X concentration and then diluted with sterile water and various NSAIDs (ibuprofen, salicylic acid, acetylsalicylic acid) or acetaminophen (Sigma, St. Louis). Overnight cultures grown with and without NSAID were diluted to 105-106 cfu/ml., A 3-cm segment of Silastic (French 2-8; Gesco International, San
110 1992; 166 (October)
Farber and Wolff
862
Table 1. Production of S. epidermidis extracellular slime to polypropylene in the presence of increasing concentrations of salicylic acid. Concentration (mM)
OD
Weight (mg/L)
TSB control I 2 5 10 25
1.5 1.4 1.3 0.5 0.08 0.01
83 66 NO 63 ND
42
NOTE. Optical density (mean offive trials) was determined at 570 nm. TSB. trypticase soy broth; ND. not determined.
Antonio, TX) or polyurethane (Becton Dickinson, Dayton, OH) catheter tubing on a 2 I-gauge needle was placed into the culture medium for IS min at room temperature. The catheter was washed three times in sterile saline by agitation and by drawing saline through the needle three times with a 3-mL syringe fitted to the needle. A section (I em) of the proximal catheter was cut off and discarded; the remaining 2-cm section was rolled over a blood agar plate in four directions, plates were incubated overnight, and colonies were counted as described [6]. The observer was blinded during the counting process. Incorporation of NSAID into polymers. The NSAIDs were incorporated into the catheter segments by heating and swelling [11]. The catheter segments were prewarmed in a 68°C water bath overnight and air dried the next day. They were then immersed into NSAIDs in 95% ethanol at -20°C for 2 h and air dried. The coated catheters were then placed into S. epidermidis cultures in TSB at 37°C for 4 h unless otherwise stated. They were then washed and quantitatively cultured as previously described. Bioluminescent assay for bacterial attachment. Bacterial attachment to the polymers and polypropylene wells was measured by modification of the method of Ludwicka et al. [12]. Catheter segments (5 mm) that had been pretreated with NSAID or ethanol alone were placed into S. epidermidis cultures for varying times and at different temperatures. The segments were vigorously washed with normal saline and placed into microtiterwells (Dynatech) to which 100 uL of'luciferin-luciferase (Analytical Luminescence Lab, San Diego) was added to release ATP. ATP-releasing agent (50 JLL) was then added to release AMP, pyrophosphate, and light. The light emission was read on a luminometer (Dynatech). A stock ATP solution (Analytical Luminescence) was diluted, and light units were plotted to establish a standard curve. Bacteria were also grown directly in the microtiter wells with varying concentrations of NSAID. The cultures were gently aspirated and the wells washed with saline. The luciferase and ATP-releasing agent were added as previously described, and the wells were read in the luminometer. Radiolabeling ofbacteria. S. epidermidis was radiolabeled by including I JLCi of[ 14C]sodium acetate (NEN, Boston) into the overnight TSB culture [6]. Catheters pretreated with NSAIDs or ethanol alone were incubated in the overnight culture (10 8 cfu/ mL) for 4 h at 37°C. Control catheters that had been similarly
treated were incubated in sterile TSB with 1 JLCi of[ 14C]sodium acetate for 4 h at 37°C. After washing three times, the catheter segments were dried and counted in scintillation vials. All studies were done in triplicate with the results expressed as the mean of three trials. Analysis of data. Analysis of variance for significant differences was done across groups using the Dunnett's test for comparing conditions to controls.
Results Table I shows the production of extracellular slime in increasing concentrations ofsalicylic acid. Both optical density and actual weight measurements demonstrated decreasing amounts of slime production in the presence of increasing concentrations of salicylic acid. The effects of salicylic acid, ibuprofen, acetylsalicylic acid, and acetaminophen on the growth characteristics of S. epidermidis were studied in chemically defined mediums; there was minimal inhibition of growth except with ibuprofen (table 2). Table 3 shows the adherence ofcoagulase-negative staphylococci to various polymers in increasing concentrations of salicylic acid. There was a progressive decrease in adherence as measured by the rolling assay for all of the polymers studied (Silastic, polyvinyl chloride, polyurethane, and Teflon). This decrease was significant (P < .006). In trials evaluating the adherence of S. epidermidis grown in the presence of 2 mM NSAID to Silastic catheters, a highly significant inhibition ofadherence was seen with all NSAIDs except acetylsalicylic acid (P < .000 I). No significant inhibition was observed with acetaminophen (table 4). In a similar study of the adherence of S. epidermidis to polyurethane catheter segments in the presence of 2 mM NSAID, significant differences were observed between acetaminophen and the NSAIDs except phenylbutazone (table 5). The differences between control and NSAIDs (except phenylbutazone) were significant (P < .008). Table 6 demonstrates the adherence of S. epidermidis to polyurethane catheters that were treated with 500 mM NSAID. Salicylic acid completely inhibited adherence, while ibuprofen and acetaminophen had no effect. The difference between salicylic
Table 2. Growth of S. epidermidis in increasing concentrations of nonsteroidal antiinflammatory agents. Growth (cfu) at concentration
Ethanol control Salicylic acid Acetaminophen Acetylsalicylic acid Ibuprofen
ImM
2mM
5mM
10mM
2.0 X 109 3.8 X 109 3.0 X 109
2.2 X 109 2.2 X 109 1-2 X 109
1.3 X 109 1.1 X 109 1.5 X 109
6.4 X 108 4.0 X 107 4.8 X 108
2.4 X 109 1.2 X 109
1-4 X 109 6.0 X 108
1.1 X 109 6.8 X 107
2.4 X 108 3.6 X 106
Table 3. Adherence of S. epidermidis grown in increasing concentrations of salicylic acid to various medical polymers.
Polymer. concentration Silastic Control
Polyvinyl chloride Control
ImM 5mM
50.3
ImM 5mM
89.6 27.6 16.2
ImM 5mM
Adherence (cfu/plate)
24 52
41 43
Control Salicylic acid Acetylsalicylic acid Ibuprofen Acetaminophen Indomethacin Phenylbutazone NOTE.
130.8 129.8
32
Teflon Control
Table 5. Adherence of S. epidermidis grown in various nonsteroidal antiinflammatory drugs to polyurethane catheters.
% inhibition
84.7 48
Polyurethane Control
NOTE.
Adherence (cfu/plate)
69.8 53.2 33.3
ImM 5mM
863
Salicylic Acid Coating of Medical Polymers
lID 1992; 166 (October)
% inhibition
92 36 66 64
30
107 7
92
86
6
61 28
o
Results are means of three trials.
I 75
69 82
Results are means of five trials.
acid and the other compounds was significant (P < .000 I). A similar study was done with polyurethane catheter segments that had been pretreated in 200 or 600 mM salicylic acid. The catheters were incubated with S. epidermidis (4 X 103 cfu/ml.) for 4 h at 37°C in TSB. Inhibition of adherence was 75% with 200 mM salicylic acid and 78% with 600 mM salicylic acid. Table 7 lists S. epidermidis adherence as measured by bioluminescent and spectrophotometric methods with organisms grown directly in polypropylene wells for 24 h in various concentrations of NSAID. As measured by both methods, there was a decrease in adherence out of proportion to the modest decrease in growth of the organisms; this decrease was seen with ibuprofen, acetylsalicylic acid, and salicylic acid. No change was noted with 10 mM acetaminophen by bioluminescence and only a minimal decrease was noted by the spectrophotometric assay. Another trial (table
8) was done by incubating polyurethane tubing that had been soaked in various concentrations ofNSAID with bacteria and then determining adherence by bioluminescence. Bacteria were less adherent to polyurethane that had been incubated in NSAID. No effect was seen with acetaminophen. Table 9 lists the results of incubating Silastic catheters in medium containing radiolabeled S. epidermidis. Nonspecific binding was observed between salicylate-coated catheters and [ 14C]sodium acetate. Therefore, adjusted counts per minute were calculated for each paired sample by subtracting the number in a simultaneously run sample containing no bacteria. A progressive decrease in adherence was seen in catheters that had been soaked in 200 and 800 mM salicylic acid. Table 10 gives results of an agar rolling study in which polyurethane catheters were pretreated with 200 mM salicylic acid or ibuprofen. They were then incubated in 5 X 105 cfu/ mL S. epidermidis for 4 h at 37°C in 20% horse serum. Both salicylic acid- and ibuprofen-treated catheters were resistant to bacterial adherence.
Discussion In recent years, coagulase-negative staphylococci have emerged as the most important cause of medical device-related infection [I]. The pathophysiology of these infections
Table 4. Adherence ofS. epidermidis grown in various nonsteroidal antiinflammatory drugs to Silastic catheters. Adherence (cfu/plate) Control Salicylic acid Acetylsalycylic acid Acetaminophen Ibuprofen Indomethacin Phenylbutazone NOTE.
869 519
853 1085 611 154
215
Results are means of three trials.
Table 6. Adherence of S. epidermidis to treated polyurethane catheters. % inhibition
Adherence (cfu/plate)
40 2 25 30 82 75
Control Salicylic acid Acetylsalicylic acid Ibuprofen Acetaminophen NOTE.
% inhibition
44
o
100
16 71
64
72
Results are means of three trials.
o o
Farber and Wolff
864
Table 7. Bioluminescent assay of adherence of S. epidermidis grown in polypropylene wells.
Treatment concentration Control (ethanol) Ibuprofen ImM 2mM 5mM 10mM Acetylsalicylate ImM 2mM 5mM 10mM Salicylic acid ImM 2mM 5mM 10mM Acetaminophen ImM 2mM 5mM 10mM NOTE.
Bacterial ATP (pmol)
OD
9 X 10- 9
0.83
9.5 5.0 2.0 8.5
X 10- 9
X 10- 11
0.596 0.540 0.134 0.065
>10- 9 2 X 10- 9 2 X 10- 11 4.5 X 10- 11
0.490 0.911 0.575 0.428
2 4 7 2
X 10- 10 X 10- 11
X X X X
10- 10 10- 10 10- 10 10- 11
0.503 0.556 0.117 0.046
>10- 9 1.5 X 10-9 1.1 X 10- 10 7.0 X 10- 9
1.33 1.47 0.93 0.29
Optical density was determined at 570 nm.
is complex and is related to properties of the host, the device, and the organism. The marked propensity ofcoagulase-negative staphylococci to adhere to polymers appears to be related to capsular and extracellular polysaccharide production [1-5]. We studied the effect of NSAIDs on adherence of S. epidermidis to medical polymers. Salicylic acid did not significantly inhibit the growth of this organism. Nevertheless, it did decrease the adherence as measured directly and by a spectrophotometric assay. Similar results were obtained with ibuprofen. In addition, organisms that were grown in salicylate were less likely to adhere to Silastic, polyurethane, Tef-
Table 8. Bioluminescent assay ofS. epidermidis adherence to nonsteroidal antiinflammatory drug-treated Silastic and polyurethane catheters. Treatment Control Salicylic acid Salicylic acid Ibuprofen Ibuprofen Acetaminophen Indomethacin Acetylsalicylate NOTE.
Concen tration
Silastic
Polyurethane
200mM 800mM 200mM 400mM 200mM 100mM 200mM
8 X 10- 11 5 X 10- 13 9 X 10- 13 2 X 10- 12 3 X 10- 13 9 X 10- 11 7 X 10- 1 1 I X 10- 14
4 X 10- 11 9 X 10- 12 6 X 10- 12 I X 10- 14 I X 10- 13 3 X 10- 11 I X 10- 11 9 X 10- 13
Results are pmol of bacterial ATP, mean of three trials.
1ID 1992; 166 (October)
Table 9. Adherence of radiolabeled S. epidermidis to salicylic acid-treated Silastic tubing as measured by radiolabeling assay.
cpm
Adjusted cpm
35.8 146.5
110.7
87.8 176.9
89.1
236.5 242.1
5.6
Pretreatment, incubation Ethanol No bacteria Bacteria Salicylic acid, 200 mM No bacteria Bacteria Salicylic acid, 800 mM No bacteria Bacteria NOTE.
Results are means of nine trials.
lon, and polyvinyl chloride catheters, as shown by quantification using a rolling assay. After demonstrating that the addition of salicylic acid to medium could decrease the adherence of organisms to polymers, we attempted to coat the polymers with this and other NSAIDs. Previous studies have demonstrated that soaking polymers in a cold ethanol solution can promote the absorption of a number of compounds into selected polymers [II]. Three separate methods were used to evaluate the ability of treated polymers to resist adherence. Rolling the catheters on agar plates, radiolabeling the bacteria, and using a bioluminescence technique all produced consistent results, although not all materials were studied. The one exception was with Teflon catheters. It is well established that Teflon cannot be coated, and no decrease in adherence was noted when these catheters were soaked in the NSAID. A variety of attempts have been made to decrease the adherence of bacteria to catheters. These have included antibiotic coating of catheters, the use of silver-impregnated collagen cuffs, and the use of chlorhexidine-coated catheters [13-15]. In addition, in vitro studies have suggested that a number of properties directly related to the polymer itself may affect bacterial adherence [16]. We believe that the use ofNSAIDs to coat medical devices may be a novel approach in attempting to decrease the incidence of infection. Unlike antimicrobial coating, the current approach does not rely on the inhibition of bacterial growth.
Table 10. Adherence of S. epidermidis to pretreated polyurethane catheters in horse serum,
Ethanol Salicylic acid Ibuprofen
Adherence (cfu/rnm)
% inhibition
11.0 1.9 0.4
83 97
JID 1992; 166 (October)
Salicylic Acid Coating of Medical Polymers
NSAIDs probably decrease adherence by inhibiting slime and perhaps capsular polysaccharide production. The exact mechanism by which these compounds decrease slime production is unknown. Many studies demonstrate that these compounds have a number of actions on cell membranes independent of their inhibition of prostagladins [17]. These actions are not seen with acetaminophen, which did not inhibit adherence in our study. In addition to antiinfective properties, it is possible that NSAIDs might decrease the incidence of thrombophlebitis. Salicylic acid has significantly less activity in inhibiting platelet thromboxane than does acetylsalicylic acid (aspirin) and is a nontoxic metabolite. Plasma levels of22 nmoljmL have been measured after administration of 325 mg of aspirin [18]. Nevertheless, all of the NSAIDs inhibit platelet aggregation [17]. Many issues regarding the effect of the NSAIDs on bacterial adherence remain to be resolved. In addition to the mechanism, the extent of this interaction with other organisms needs to be determined. We treated polymers by simply soaking them in the NSAID in ethanol. Other delivery methods, such as the use of tridodecylemethylammonium chloride, have been successfully used to provide a slow-release system with antibiotic-coated polymers [19]. These and other coating methods and their kinetics need to be explored. There have been no previous reports describing the potential use of NSAIDs in preventing medical device-related infection. However, a recent report suggests that perioperative aspirin protects against early total prosthetic joint infection [20]. The prophylactic use of systemic NSAIDs before insertion of medical devices also deserves further attention. Acknowledgments
We appreciate the help of M. Lesser for statistical analysis, Ellen Sweeney for preparing the manuscript, and David Shepp for critical review.
References I. Archer GL. Staphylococcus epidennidis: the organism, its diseases, and treatment. Cu~ Clin Top Infect Dis 1984;7:25-46.' 2. Davenport DS, Massanari RM, Pfaller MA, Bale MJ, Streed SA, Hierholzer WJ J r. Usefulness ofa test for slime production as a marker for clinically significant infections with coagulase-negative staphylococci. J Infect Dis 1986; 153:332-9. 3. Christensen GD, Parisi JT, Bisno AL, Simpson WA, Beachey EH. Characterization of clinically significant strains of coagulase-negative staphylococci. J Clin Microbiol 1983; 18:258-69.
865
4. Diaz-Mitoma F, Harding GKM, Hoban DJ, Roberts RS, Low DE. Clinical significance ofa test for slime production in ventriculoperitoneal shunt infections caused by coagulase-negative staphylococci. J Infect Dis 1987;156:555-60. 5. Christensen GD, Barker LP, Mawhinney TP, Baddour LM, Simpson WA. Identification of an antigenic marker of slime production for Staphylococcus epidermidis. Infect Immun 1990;58:2906-11. 6. Tojo M, Yamashita N, Goldmann DA, PierGB. Isolation and characterization ofa capsular polysaccharide adhesin from Staphylococcus epidermidis. J Infect Dis 1988;157:713-22. 7. Hussain M, Hastings JGM, White PJ. Isolation and composition of the extracellular slime made by coagulase-negative staphylococci in chemically defined medium. J Infect Dis 1991; 163:534-41. 8. Domenico P, Schwartz S, Cunha BA. Reduction ofcapsular polysaccharide production in Klebsiella pneumoniae by sodium salicylate. Infect Immun 1989;57:3778-82. 9. Farber BF, Kaplan MH. Clogston AG. Staphylococcus epidermidis extracted slime inhibits the antimicrobial action of glycopeptide antibiotics. J Infect Dis 1990;161:37-40. 10. Christensen GD, Simpson WA, Younger JJ, et al. Adherence of coagulase negative staphylococci to plastic tissue culture plates: a quantitative model for the adherence of staphylococci to medical devices. J Clin Microbiol 1985;22:996-1006. II. Jansen B, Schareina S, Steinhauser H, Peters G, Schumacher-Perdreau F, Pulverer G. Development of polymers with anti-infectious properties. Polymeric Material Sci Eng 1987;57:43-52. 12. Ludwicka A, Switalski LM, Lundin A, Pulverer G, Wadstrom T. Bioluminescent assay for measurement of bacterial attachment to polyethylene. J Microbiol Methods 1985;4: 169-77. 13. Kamal GD, Pfaller MA, Rempe LE, Jebson PJR. Reduced intravascular catheter infection by antibiotic bonding. JAMA 1991;265:23648. 14. Flowers RH III, Schwenzer KJ, Kopel RF, Fisch MJ, Tucker SI, Farr BM. Efficacy of an attachable subcutaneous cuff for the prevention of intravascular catheter-related infection: a randomized controlled trial. JAMA 1989;261 :878-83. 15. Maki DG, Wheeler SJ, Stolz SM, Mermel LA. Clinical trial ofa novel antiseptic catheter [abstract 461]. In: Program and abstracts of the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, DC: American Society for Microbiology, 1991. 16. Sawant AD, Gabriel M, Mayo MB, Ahearn DG. Radiopacity additives in silicone stent materials reduce in vitro bacterial adherence. Curr Microbiol 1991 ;22:285-92. 17. Abramson SB, Weissman G. The mechanisms of action of non-steroidal anti-inflammatory drugs. Arthritis Rheum 1989;32: 1-9. 18. Clarke RJ, Mayo G, Price P, Fitzgerald GA. Suppression ofthromboxane A 2 but not of systemic prostacyclin by controlled release aspirin. N Engl J Med 1991 ;325: 1137-41. 19. Trooskin SZ, Donetz AP, Baxter J, Harvey RA, Greco RS. Infection resistant continuous peritoneal dialysis catheters. Nephron 1987;46:263-7. 20. Wordnick CL, Hickingbotham N, Pasternak H, Maderazo EG. Postoperative aspirin and protection following joint replacement surgery [abstract 1283]. In: Programs and abstracts of the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, DC: American Society for Microbiology, 1991.
The Use of Nonsteroidal Antiinflammatory Drugs to Prevent Adherence of Staphylococcus epidermidis to Medical Polymers Bruce F. Farber and Arlene G. Wolff
Division of Infectious Diseases. Department of Medicine. North Shore University Hospital-Cornell University Medical College, Manhasset. New York
The effect of salicylates and other nonsteroidal antiinflammatory drugs on the production of Staphylococcus epidermidis extracellular slime was studied. A dose-related decrease in slime production was observed with increasing concentrations of salicylic acid. S. epidermidis grown in 5 mM salicylic acid were less likely to adhere to Silastic, polyvinyl chloride, polyurethane, and Teflon catheters (P < .006); strains grown in 2 mM salicylic acid, ibuprofen, indomethacin, or phenylbutazone were less adherent to Silastic catheters (P < .001). Similar results were obtained with polyurethane catheters. S. epidermidis strains were less likely to adhere (43%-82% inhibition) to polyurethane catheters treated with 500 mM salicylic acid diluted in ethanol (P < .0001). Similar differences were not observed with acetaminophen, ibuprofen, or acetylsalicylic acid. Adherence of radiolabeled S. epidermidis to salicylic acid-treated Silastic catheters demonstrated a dose-related reduction. The use of salicylic acid to coat medical devices may decrease the incidence of device-related infection. Staphylococcus epidermidis has emerged as a major pathogen over the past decade. Previously considered contaminants, these organisms are the most common cause of prosthetic device- and catheter-related infection [1]. Certain strains of S. epidermidis produce an extracellular slime-like substance that appears to be an important virulence factor [2-4]. Recent studies have characterized the composition of S. epidermidis slime and its role in promoting the adherence of bacteria to polymers [5-7]. Extracellular slime produced by S. epidermidis is predominantly made of polysaccharides rich in glucose and galactose [6, 7]. Domenico et a1. [8] demonstrated that sodium salicylate reduces the amount of capsular polysaccharide produced by Klebsiella pneumoniae. Because S. epidermidis extracellular slime is a form ofexuberent capsular polysaccharide, we studied the effect of salicylates on slime production. We also studied the effect ofcoating medical polymers with nonsteroidal antiinflammatory drugs (NSAIDs) on bacterial adherence. These observations may have utility in the prevention of device-related infections.
Methods Preparation of extracts. Coagulase-negative staphylococci previously identified as S. epidermidis [9] were used in these
Received 18 February 1992; revised 4 May 1992. Financial support: Jane and Dayton Brown and Dayton Brown Jr. family fund. Reprints or correspondence: Dr. Bruce F. Farber, North Shore University Hospital-Cornell University Medical College, 300 Community Dr., Manhasset. NY I 1030.
The Journal of Infectious Diseases 1992;166:861-5 © 1992 by The University of Chicago. All rights reserved. 0022-1899/92/6604-0025$01.00
experiments. All strains had been found to be slime-positive by a qualitative method [10]. Crude slime filtrates were prepared from strains grown in chemically defined medium supplemented with various concentrations of sodium salicylate. The cultures were incubated for 48 h with shaking at 37°C and were centrifuged for 30 min at 27,000 g. The supernatant was dialyzed against sterile water for 48 h and then lyophilized. The freeze-dried product was reconstituted in 1/ I0 saturated sodium acetate, subjected to phenol and chloroform extractions as previously described, lyophilized, and weighed [9]. Spectrophotometric method. Adherence of coagulase-negative staphylococci to plastic culture plates was quantitatively determined by the method ofChristensen et a1. [10]. Organisms were grown in trypticase soy broth (TSB) for 18 h at 37°C. The cultures were then diluted I: I00 with fresh TSB with and without various concentrations of sodium salicylate. Aliquots (0.2 mL) of the diluted culture were pipetted into individual wells of sterile, polystyrene, 96-well flat-bottomed tissue culture plates (Corning Science Products, Medfield, MA). The plates were incubated for 18 h at 37°C. The contents of each well were gently aspirated and the well was washed four times with 0.2 mL of PBS. Adherent organisms were fixed with Bouin's fixative and then stained with Hucker crystal violet. Excess stain was rinsed off by placing the plate under running water. After drying. the optical densities of the stained adherent bacterial films were read (Micro ELISA Auto Reader, MR 600; Dynatech, Alexandria, VA) using a wavelength of 570 nm on the single-wavelength mode. Colony counts were also done from broth cultures in the presence ofincreasing concentrations ofNSAIDs. The cultures were grown in TSB. Adherence ofbacteria to catheter segments. TSB was prepared at 2X concentration and then diluted with sterile water and various NSAIDs (ibuprofen, salicylic acid, acetylsalicylic acid) or acetaminophen (Sigma, St. Louis). Overnight cultures grown with and without NSAID were diluted to 105-106 cfu/ml., A 3-cm segment of Silastic (French 2-8; Gesco International, San
110 1992; 166 (October)
Farber and Wolff
862
Table 1. Production of S. epidermidis extracellular slime to polypropylene in the presence of increasing concentrations of salicylic acid. Concentration (mM)
OD
Weight (mg/L)
TSB control I 2 5 10 25
1.5 1.4 1.3 0.5 0.08 0.01
83 66 NO 63 ND
42
NOTE. Optical density (mean offive trials) was determined at 570 nm. TSB. trypticase soy broth; ND. not determined.
Antonio, TX) or polyurethane (Becton Dickinson, Dayton, OH) catheter tubing on a 2 I-gauge needle was placed into the culture medium for IS min at room temperature. The catheter was washed three times in sterile saline by agitation and by drawing saline through the needle three times with a 3-mL syringe fitted to the needle. A section (I em) of the proximal catheter was cut off and discarded; the remaining 2-cm section was rolled over a blood agar plate in four directions, plates were incubated overnight, and colonies were counted as described [6]. The observer was blinded during the counting process. Incorporation of NSAID into polymers. The NSAIDs were incorporated into the catheter segments by heating and swelling [11]. The catheter segments were prewarmed in a 68°C water bath overnight and air dried the next day. They were then immersed into NSAIDs in 95% ethanol at -20°C for 2 h and air dried. The coated catheters were then placed into S. epidermidis cultures in TSB at 37°C for 4 h unless otherwise stated. They were then washed and quantitatively cultured as previously described. Bioluminescent assay for bacterial attachment. Bacterial attachment to the polymers and polypropylene wells was measured by modification of the method of Ludwicka et al. [12]. Catheter segments (5 mm) that had been pretreated with NSAID or ethanol alone were placed into S. epidermidis cultures for varying times and at different temperatures. The segments were vigorously washed with normal saline and placed into microtiterwells (Dynatech) to which 100 uL of'luciferin-luciferase (Analytical Luminescence Lab, San Diego) was added to release ATP. ATP-releasing agent (50 JLL) was then added to release AMP, pyrophosphate, and light. The light emission was read on a luminometer (Dynatech). A stock ATP solution (Analytical Luminescence) was diluted, and light units were plotted to establish a standard curve. Bacteria were also grown directly in the microtiter wells with varying concentrations of NSAID. The cultures were gently aspirated and the wells washed with saline. The luciferase and ATP-releasing agent were added as previously described, and the wells were read in the luminometer. Radiolabeling ofbacteria. S. epidermidis was radiolabeled by including I JLCi of[ 14C]sodium acetate (NEN, Boston) into the overnight TSB culture [6]. Catheters pretreated with NSAIDs or ethanol alone were incubated in the overnight culture (10 8 cfu/ mL) for 4 h at 37°C. Control catheters that had been similarly
treated were incubated in sterile TSB with 1 JLCi of[ 14C]sodium acetate for 4 h at 37°C. After washing three times, the catheter segments were dried and counted in scintillation vials. All studies were done in triplicate with the results expressed as the mean of three trials. Analysis of data. Analysis of variance for significant differences was done across groups using the Dunnett's test for comparing conditions to controls.
Results Table I shows the production of extracellular slime in increasing concentrations ofsalicylic acid. Both optical density and actual weight measurements demonstrated decreasing amounts of slime production in the presence of increasing concentrations of salicylic acid. The effects of salicylic acid, ibuprofen, acetylsalicylic acid, and acetaminophen on the growth characteristics of S. epidermidis were studied in chemically defined mediums; there was minimal inhibition of growth except with ibuprofen (table 2). Table 3 shows the adherence ofcoagulase-negative staphylococci to various polymers in increasing concentrations of salicylic acid. There was a progressive decrease in adherence as measured by the rolling assay for all of the polymers studied (Silastic, polyvinyl chloride, polyurethane, and Teflon). This decrease was significant (P < .006). In trials evaluating the adherence of S. epidermidis grown in the presence of 2 mM NSAID to Silastic catheters, a highly significant inhibition ofadherence was seen with all NSAIDs except acetylsalicylic acid (P < .000 I). No significant inhibition was observed with acetaminophen (table 4). In a similar study of the adherence of S. epidermidis to polyurethane catheter segments in the presence of 2 mM NSAID, significant differences were observed between acetaminophen and the NSAIDs except phenylbutazone (table 5). The differences between control and NSAIDs (except phenylbutazone) were significant (P < .008). Table 6 demonstrates the adherence of S. epidermidis to polyurethane catheters that were treated with 500 mM NSAID. Salicylic acid completely inhibited adherence, while ibuprofen and acetaminophen had no effect. The difference between salicylic
Table 2. Growth of S. epidermidis in increasing concentrations of nonsteroidal antiinflammatory agents. Growth (cfu) at concentration
Ethanol control Salicylic acid Acetaminophen Acetylsalicylic acid Ibuprofen
ImM
2mM
5mM
10mM
2.0 X 109 3.8 X 109 3.0 X 109
2.2 X 109 2.2 X 109 1-2 X 109
1.3 X 109 1.1 X 109 1.5 X 109
6.4 X 108 4.0 X 107 4.8 X 108
2.4 X 109 1.2 X 109
1-4 X 109 6.0 X 108
1.1 X 109 6.8 X 107
2.4 X 108 3.6 X 106
Table 3. Adherence of S. epidermidis grown in increasing concentrations of salicylic acid to various medical polymers.
Polymer. concentration Silastic Control
Polyvinyl chloride Control
ImM 5mM
50.3
ImM 5mM
89.6 27.6 16.2
ImM 5mM
Adherence (cfu/plate)
24 52
41 43
Control Salicylic acid Acetylsalicylic acid Ibuprofen Acetaminophen Indomethacin Phenylbutazone NOTE.
130.8 129.8
32
Teflon Control
Table 5. Adherence of S. epidermidis grown in various nonsteroidal antiinflammatory drugs to polyurethane catheters.
% inhibition
84.7 48
Polyurethane Control
NOTE.
Adherence (cfu/plate)
69.8 53.2 33.3
ImM 5mM
863
Salicylic Acid Coating of Medical Polymers
lID 1992; 166 (October)
% inhibition
92 36 66 64
30
107 7
92
86
6
61 28
o
Results are means of three trials.
I 75
69 82
Results are means of five trials.
acid and the other compounds was significant (P < .000 I). A similar study was done with polyurethane catheter segments that had been pretreated in 200 or 600 mM salicylic acid. The catheters were incubated with S. epidermidis (4 X 103 cfu/ml.) for 4 h at 37°C in TSB. Inhibition of adherence was 75% with 200 mM salicylic acid and 78% with 600 mM salicylic acid. Table 7 lists S. epidermidis adherence as measured by bioluminescent and spectrophotometric methods with organisms grown directly in polypropylene wells for 24 h in various concentrations of NSAID. As measured by both methods, there was a decrease in adherence out of proportion to the modest decrease in growth of the organisms; this decrease was seen with ibuprofen, acetylsalicylic acid, and salicylic acid. No change was noted with 10 mM acetaminophen by bioluminescence and only a minimal decrease was noted by the spectrophotometric assay. Another trial (table
8) was done by incubating polyurethane tubing that had been soaked in various concentrations ofNSAID with bacteria and then determining adherence by bioluminescence. Bacteria were less adherent to polyurethane that had been incubated in NSAID. No effect was seen with acetaminophen. Table 9 lists the results of incubating Silastic catheters in medium containing radiolabeled S. epidermidis. Nonspecific binding was observed between salicylate-coated catheters and [ 14C]sodium acetate. Therefore, adjusted counts per minute were calculated for each paired sample by subtracting the number in a simultaneously run sample containing no bacteria. A progressive decrease in adherence was seen in catheters that had been soaked in 200 and 800 mM salicylic acid. Table 10 gives results of an agar rolling study in which polyurethane catheters were pretreated with 200 mM salicylic acid or ibuprofen. They were then incubated in 5 X 105 cfu/ mL S. epidermidis for 4 h at 37°C in 20% horse serum. Both salicylic acid- and ibuprofen-treated catheters were resistant to bacterial adherence.
Discussion In recent years, coagulase-negative staphylococci have emerged as the most important cause of medical device-related infection [I]. The pathophysiology of these infections
Table 4. Adherence ofS. epidermidis grown in various nonsteroidal antiinflammatory drugs to Silastic catheters. Adherence (cfu/plate) Control Salicylic acid Acetylsalycylic acid Acetaminophen Ibuprofen Indomethacin Phenylbutazone NOTE.
869 519
853 1085 611 154
215
Results are means of three trials.
Table 6. Adherence of S. epidermidis to treated polyurethane catheters. % inhibition
Adherence (cfu/plate)
40 2 25 30 82 75
Control Salicylic acid Acetylsalicylic acid Ibuprofen Acetaminophen NOTE.
% inhibition
44
o
100
16 71
64
72
Results are means of three trials.
o o
Farber and Wolff
864
Table 7. Bioluminescent assay of adherence of S. epidermidis grown in polypropylene wells.
Treatment concentration Control (ethanol) Ibuprofen ImM 2mM 5mM 10mM Acetylsalicylate ImM 2mM 5mM 10mM Salicylic acid ImM 2mM 5mM 10mM Acetaminophen ImM 2mM 5mM 10mM NOTE.
Bacterial ATP (pmol)
OD
9 X 10- 9
0.83
9.5 5.0 2.0 8.5
X 10- 9
X 10- 11
0.596 0.540 0.134 0.065
>10- 9 2 X 10- 9 2 X 10- 11 4.5 X 10- 11
0.490 0.911 0.575 0.428
2 4 7 2
X 10- 10 X 10- 11
X X X X
10- 10 10- 10 10- 10 10- 11
0.503 0.556 0.117 0.046
>10- 9 1.5 X 10-9 1.1 X 10- 10 7.0 X 10- 9
1.33 1.47 0.93 0.29
Optical density was determined at 570 nm.
is complex and is related to properties of the host, the device, and the organism. The marked propensity ofcoagulase-negative staphylococci to adhere to polymers appears to be related to capsular and extracellular polysaccharide production [1-5]. We studied the effect of NSAIDs on adherence of S. epidermidis to medical polymers. Salicylic acid did not significantly inhibit the growth of this organism. Nevertheless, it did decrease the adherence as measured directly and by a spectrophotometric assay. Similar results were obtained with ibuprofen. In addition, organisms that were grown in salicylate were less likely to adhere to Silastic, polyurethane, Tef-
Table 8. Bioluminescent assay ofS. epidermidis adherence to nonsteroidal antiinflammatory drug-treated Silastic and polyurethane catheters. Treatment Control Salicylic acid Salicylic acid Ibuprofen Ibuprofen Acetaminophen Indomethacin Acetylsalicylate NOTE.
Concen tration
Silastic
Polyurethane
200mM 800mM 200mM 400mM 200mM 100mM 200mM
8 X 10- 11 5 X 10- 13 9 X 10- 13 2 X 10- 12 3 X 10- 13 9 X 10- 11 7 X 10- 1 1 I X 10- 14
4 X 10- 11 9 X 10- 12 6 X 10- 12 I X 10- 14 I X 10- 13 3 X 10- 11 I X 10- 11 9 X 10- 13
Results are pmol of bacterial ATP, mean of three trials.
1ID 1992; 166 (October)
Table 9. Adherence of radiolabeled S. epidermidis to salicylic acid-treated Silastic tubing as measured by radiolabeling assay.
cpm
Adjusted cpm
35.8 146.5
110.7
87.8 176.9
89.1
236.5 242.1
5.6
Pretreatment, incubation Ethanol No bacteria Bacteria Salicylic acid, 200 mM No bacteria Bacteria Salicylic acid, 800 mM No bacteria Bacteria NOTE.
Results are means of nine trials.
lon, and polyvinyl chloride catheters, as shown by quantification using a rolling assay. After demonstrating that the addition of salicylic acid to medium could decrease the adherence of organisms to polymers, we attempted to coat the polymers with this and other NSAIDs. Previous studies have demonstrated that soaking polymers in a cold ethanol solution can promote the absorption of a number of compounds into selected polymers [II]. Three separate methods were used to evaluate the ability of treated polymers to resist adherence. Rolling the catheters on agar plates, radiolabeling the bacteria, and using a bioluminescence technique all produced consistent results, although not all materials were studied. The one exception was with Teflon catheters. It is well established that Teflon cannot be coated, and no decrease in adherence was noted when these catheters were soaked in the NSAID. A variety of attempts have been made to decrease the adherence of bacteria to catheters. These have included antibiotic coating of catheters, the use of silver-impregnated collagen cuffs, and the use of chlorhexidine-coated catheters [13-15]. In addition, in vitro studies have suggested that a number of properties directly related to the polymer itself may affect bacterial adherence [16]. We believe that the use ofNSAIDs to coat medical devices may be a novel approach in attempting to decrease the incidence of infection. Unlike antimicrobial coating, the current approach does not rely on the inhibition of bacterial growth.
Table 10. Adherence of S. epidermidis to pretreated polyurethane catheters in horse serum,
Ethanol Salicylic acid Ibuprofen
Adherence (cfu/rnm)
% inhibition
11.0 1.9 0.4
83 97
JID 1992; 166 (October)
Salicylic Acid Coating of Medical Polymers
NSAIDs probably decrease adherence by inhibiting slime and perhaps capsular polysaccharide production. The exact mechanism by which these compounds decrease slime production is unknown. Many studies demonstrate that these compounds have a number of actions on cell membranes independent of their inhibition of prostagladins [17]. These actions are not seen with acetaminophen, which did not inhibit adherence in our study. In addition to antiinfective properties, it is possible that NSAIDs might decrease the incidence of thrombophlebitis. Salicylic acid has significantly less activity in inhibiting platelet thromboxane than does acetylsalicylic acid (aspirin) and is a nontoxic metabolite. Plasma levels of22 nmoljmL have been measured after administration of 325 mg of aspirin [18]. Nevertheless, all of the NSAIDs inhibit platelet aggregation [17]. Many issues regarding the effect of the NSAIDs on bacterial adherence remain to be resolved. In addition to the mechanism, the extent of this interaction with other organisms needs to be determined. We treated polymers by simply soaking them in the NSAID in ethanol. Other delivery methods, such as the use of tridodecylemethylammonium chloride, have been successfully used to provide a slow-release system with antibiotic-coated polymers [19]. These and other coating methods and their kinetics need to be explored. There have been no previous reports describing the potential use of NSAIDs in preventing medical device-related infection. However, a recent report suggests that perioperative aspirin protects against early total prosthetic joint infection [20]. The prophylactic use of systemic NSAIDs before insertion of medical devices also deserves further attention. Acknowledgments
We appreciate the help of M. Lesser for statistical analysis, Ellen Sweeney for preparing the manuscript, and David Shepp for critical review.
References I. Archer GL. Staphylococcus epidennidis: the organism, its diseases, and treatment. Cu~ Clin Top Infect Dis 1984;7:25-46.' 2. Davenport DS, Massanari RM, Pfaller MA, Bale MJ, Streed SA, Hierholzer WJ J r. Usefulness ofa test for slime production as a marker for clinically significant infections with coagulase-negative staphylococci. J Infect Dis 1986; 153:332-9. 3. Christensen GD, Parisi JT, Bisno AL, Simpson WA, Beachey EH. Characterization of clinically significant strains of coagulase-negative staphylococci. J Clin Microbiol 1983; 18:258-69.
865
4. Diaz-Mitoma F, Harding GKM, Hoban DJ, Roberts RS, Low DE. Clinical significance ofa test for slime production in ventriculoperitoneal shunt infections caused by coagulase-negative staphylococci. J Infect Dis 1987;156:555-60. 5. Christensen GD, Barker LP, Mawhinney TP, Baddour LM, Simpson WA. Identification of an antigenic marker of slime production for Staphylococcus epidermidis. Infect Immun 1990;58:2906-11. 6. Tojo M, Yamashita N, Goldmann DA, PierGB. Isolation and characterization ofa capsular polysaccharide adhesin from Staphylococcus epidermidis. J Infect Dis 1988;157:713-22. 7. Hussain M, Hastings JGM, White PJ. Isolation and composition of the extracellular slime made by coagulase-negative staphylococci in chemically defined medium. J Infect Dis 1991; 163:534-41. 8. Domenico P, Schwartz S, Cunha BA. Reduction ofcapsular polysaccharide production in Klebsiella pneumoniae by sodium salicylate. Infect Immun 1989;57:3778-82. 9. Farber BF, Kaplan MH. Clogston AG. Staphylococcus epidermidis extracted slime inhibits the antimicrobial action of glycopeptide antibiotics. J Infect Dis 1990;161:37-40. 10. Christensen GD, Simpson WA, Younger JJ, et al. Adherence of coagulase negative staphylococci to plastic tissue culture plates: a quantitative model for the adherence of staphylococci to medical devices. J Clin Microbiol 1985;22:996-1006. II. Jansen B, Schareina S, Steinhauser H, Peters G, Schumacher-Perdreau F, Pulverer G. Development of polymers with anti-infectious properties. Polymeric Material Sci Eng 1987;57:43-52. 12. Ludwicka A, Switalski LM, Lundin A, Pulverer G, Wadstrom T. Bioluminescent assay for measurement of bacterial attachment to polyethylene. J Microbiol Methods 1985;4: 169-77. 13. Kamal GD, Pfaller MA, Rempe LE, Jebson PJR. Reduced intravascular catheter infection by antibiotic bonding. JAMA 1991;265:23648. 14. Flowers RH III, Schwenzer KJ, Kopel RF, Fisch MJ, Tucker SI, Farr BM. Efficacy of an attachable subcutaneous cuff for the prevention of intravascular catheter-related infection: a randomized controlled trial. JAMA 1989;261 :878-83. 15. Maki DG, Wheeler SJ, Stolz SM, Mermel LA. Clinical trial ofa novel antiseptic catheter [abstract 461]. In: Program and abstracts of the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, DC: American Society for Microbiology, 1991. 16. Sawant AD, Gabriel M, Mayo MB, Ahearn DG. Radiopacity additives in silicone stent materials reduce in vitro bacterial adherence. Curr Microbiol 1991 ;22:285-92. 17. Abramson SB, Weissman G. The mechanisms of action of non-steroidal anti-inflammatory drugs. Arthritis Rheum 1989;32: 1-9. 18. Clarke RJ, Mayo G, Price P, Fitzgerald GA. Suppression ofthromboxane A 2 but not of systemic prostacyclin by controlled release aspirin. N Engl J Med 1991 ;325: 1137-41. 19. Trooskin SZ, Donetz AP, Baxter J, Harvey RA, Greco RS. Infection resistant continuous peritoneal dialysis catheters. Nephron 1987;46:263-7. 20. Wordnick CL, Hickingbotham N, Pasternak H, Maderazo EG. Postoperative aspirin and protection following joint replacement surgery [abstract 1283]. In: Programs and abstracts of the 31st Interscience Conference on Antimicrobial Agents and Chemotherapy. Washington, DC: American Society for Microbiology, 1991.
