Article
Vol. 11. No. 9
789
Eur. J. Clin. Microbiol. Infect. Dis., September 1992, p. 789-796 0934-9723/92/09 0789-08 $ 3.00/0
Scanning Electron Microscopy of Bacterial Biofilms on Indwelling Bladder Catheters
L. G a n d e r t o n 1, j. Chawla 2, C. Winters 1, J. W i m p e n n y 1, D. Stickler 1.
Fifty Foley bladder catheters that had been indwelling for periods ranging from 3 to 83 days (mean 35 days) were examined for the presence of bacterial biofilm. Scanning electron microscopy on freeze-dried cross-sections and fixed, critical point-dried longitudinal sections revealed biofilm formation on the luminal surfaces of 44 of the catheters. Culture of urine samples and sonicates from catheters revealed that the prevalence of bacteriuria was less than that of catheter colonization. A wide range of nosocomiai species were found colonizing the catheters, Eschertchia coli being most often isolated. The bacterial composition of the biofilms ranged from single species to mixed communities containing up to four species. There was no relationship between the length of time that the catheter had been in situ and the extent of biofilm formation. The biofilms varied in thickness from 3 to 490/Jm and were visible as layers of bacterial cells up to about 400 cells deep, embedded in a matrix.
Infection is an all too common consequence of intravascular, peritoneal and bladder catheterization (1-3). Examination of catheters taken from these sites has revealed that they are commonly colonized by adherent layers of bacteria embedded in a matrix of extra-cellular polysaccharide (4--6). In this biofilm mode of growth the bacterial cells are more resistant to antibiotics, antiseptics and host defence mechanisms than cells growing in suspension (7-9). These characteristics may well explain why infections associated with these devices are persistent and refractory to treatment with antibacterial agents
(lO).
As part of a programme aimed at understanding the nature and extent of these resistance phenomena we have devised laboratory models of the biofilms that are found on indwelling bladder catheters. Model biofilms have been produced ranging from monolayers of single species to multi-layered films composed of mixed bacterial COmmunities and it has been observed that organisms respond differently to antibacterial agents in films of different thicknesses and composition (11-13). There is a need to obtain more information about the characteristics of the
natural catheter biofilms in order to refine our laboratory models. Previous studies have revealed that urinary catheters can become colonized by single species or mixed populations of organisms. Sparse patchy coverings or confluent thick multi-layered structures have been found in surface views of the catheter lumen (6, 14, 15). One of our laboratory model systems has the facility to produce standard films of defined thickness (16). The objective of the present study was thus to obtain information about the composition and dimensions of the natural biofilms found on indwelling catheters. The previous studies using scanning electron microscopy for examination of urinary catheter biofilms have provided surface views of the luminal films from which it has been difficult to interpret their depth and structure. We have therefore developed a simple freeze-fracturing technique to produce cross-sections Of catheters for scanning electron microscopy.
Materials and Methods Catheters. Silicone or silicone coated Foley catheters
School of Pure and Applied Biology,Universityof Wales 2w~cllegeof Cardiff, PO Box 915, Cardiff CF1 3TL, UK. elsh Spinal Injuries Unit, Rookwood Hospital, Cardiff CF5 2YN, UK,
freshly removed fl'om patients resident in the Welsh Spinal Injuries Unit at Rookwood Hospital and from patients undergoing long-term catheterization and attending a clinic at the Cardiff Royal Infirmary were transported to the laboratory for examination. Details
790
E u r . J. Clin. Microbiol. Infect. Dis.
of the length of time each catheter had been in situ and the nature of any current antibiotic therapy was also obtained together with samples of urine which had been aspirated from the catheters. Viable bacterial cell counts were performed on the urine samples after incubation for 24 h at 37 *C on CLED agar (Oxoid, UK). Characterization o f the Biofihn Communities. To characterize the bacterial communities colonizing the lumen of the catheters, sections (1 cm long) were cut from the region of the catheter within the retention balloon. The sections were rinsed once gently in 10 ml of buffer (Hanks-Hepes, pH 7.4) and then placed in 10 ml of nutrient broth (Oxoid). Disruption of the luminal biofilm was then achieved by sonieation for 5 min (Transsonic Water Bath, Camtab, UK) followed by vortex mixing for 2 min. Samples of the broth suspension were then plated out on CLED agar and incubated for 24 h at 37 *C. The resulting isolates were characterized using Cowan and Steel's tables (17) and the appropriate identification kits
(API, UK.) Scanning Electron Microscopy. Sections of catheters (1 cm in length) were taken from the region adjacent to the retention balloon on the side away from the catheter tip. These were rapidly plunged into liquid nitrogen cooled propane and transferred to liquid nitrogen. Crosssections were then produced by freeze-fracturing samples in a specially designed copper block which held the catheter and a blade in position and facilitated the production of reproducible cross-sections. The samples were then freeze-dried for 24 h at -80 *C, mounted fractured surface uppermost onto aluminium stubs, sputtered with gold and examined in a JEOL JSM5200 scanning electron microscope. To observe the nature of the surfaces of the biofilms longitudinal sections of each catheter were prepared from the region adjacent to the retention balloon. The sections were fixed in 3 % glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 1 h and then washed overnight in the phosphate buffer before being post-fixed in Millonig's phosphate buffered osmium tetroxide (1.0 %) for 1 h. The samples were dehydrated in a graded series of
aqueous ethanol solutions (30-100 %). Following dehydration the samples were critical point-dried using liquid CO2. Finally, the samples were mounted on aluminium stubs, sputtered with gold and examined in the scanning electron microscope.
Results Bacterial biofilms were clearly visible on 44 of the catheters. T h e species isolated f r o m t h e 50 paired samples of catheters and urines are listed in Table 1. E s c h e r i c h i a coli was the m o s t c o m m o n biofilm organism, being isolated f r o m 4 0 % of the catheters. Bacteriuria was less c o m m o n than biofilm colonization, a n d m a n y o r g a n i s m s present o n the biofilm w e r e not r e c o v e r e d f r o m the urine. It can be seen f r o m Table 2 t h a t 34 o f the patients w e r e not receiving antibiotics at the time o f c a t h e t e r r e m o v a l and o f this g r o u p 33 had c a t h e t e r biofilms. In five o f the six cases in which no biofilm was r e c o v e r e d the patients w e r e receiving antibiotics at the time of c a t h e t e r r e m o v a l and did n o t h a v e bacteriuria. O f the o t h e r 11 patients u n d e r g o i n g t h e r a p y w h o s e catheters w e r e c o l o n i z e d by biofilms only seven had bacteriuria (> 103 cfu/ml). M o s t o f the biofilms w e r e c o m p o s e d o f mixed bacterial c o m m u n i t i e s , single species being isolated f r o m only 25 % o f the biofilms. T h e p r o p o r tions c o m p o s e d of two, t h r e e a n d f o u r species w e r e 50 %, 20 % and 5 % respectively. Figure 1 presents the d e p t h o f biofilm in relation to the time for which each c a t h e t e r had b e e n in situ. In those cases in which a cross-section o f a
Table 1: Organisms isolated from 50 paired urine and catheter specimens. No. of positive cultures
Organisms
Escherichia coli Group D streptococci Pseudomonas aeruginosa Proteus mirabilis Klebsiella oxytoca Morganella morganii Enterobacter cloacae Staphylococcus epidermidis Providencia spp. Citrobacter diversus Candida spp. Citrobacter freundii Diphtheroids Other species
Catheter
Urine
Both
20 10 9 8 6 6 5 4 4 3 2 2 2 10
17 7 4 6 5 4 6 2 2 3 0 1 2 5
16 7 4 6 4 4 4 2 2 3 0 1 2 5'
Vol. 11, 1992
791
Table 2: Presence of catheter biofilm related to current antibiotic therapy. Antibiotic treatment
Biofilm present, urine positive* Biofilm present, urine negative No biofilm present, urine positive No biofilm present, urine negative
Yes (n = 16)
No (n = 34)
Total (n = S0)
7 4 0 5
30 3 0 1
37 7 0 6
*> 103 cfu/ml.
500,
400.
300.
E 200, .o fl]
•
•
100
•
•e o
•
0 ~ . 0
20
j
4'0
•
6'0
8b"
Time in sifu (days)
Figure 1: Relationship between the thickness of biofilms and the period of time catheters were in situ. catheter revealed a biofilm of variable thickness, the upper limit of the biofilm depth was used in this analysis. In this way Figure 1 reflects the full range of biofilm depths observed. Regression analysis revealed no significant correlation between these parameters. Examination of the extent of biofilm formation in those catheters colonized by coagulase-negative staphylococci and diphtheroid rods (Table 3) shows that despite having been in situ for a mean time of 31 days, these biofilms only had a mean depth of some 15~tm. The equivalent figures observed for catheter biofilms composed of gram-negative bacilli and the enter•cocci were 39 days and 80/am.
Figure 2: (a) Freeze-dried cross-section of an 8-day-old catheter colon~ed by a film consisting of Psettdomonas aerugbtosa (1.1 x 10~ cfu/cm2), Proteus mirabilis (1.2 x 10s cfu/cm2) and Providencia stuartii (5.3 x 107 cfu/emZ). (b) Critical point-dried preparation showing the surface of the biofilm.
792
Eur. J. Clin. Microbiol. Infect. Dis.
Table 3: Characteristics of biofilms containing coagulase negative staphylococci and diphtheroids. Organism
Biofilm depth (gm)
CNS + E. coli Diphtheroids CNS + E. coli CNS + K. oxytoca Diphtheroids + M. morganii CNS CNS CNS + P. rettgeri
3- 5 20-25 5-10 10-15 5-10 20 20-30 5
Days in situ 10 28 12 63 41 42 8 42
CNS = Coagulase-negative staphylococci.
Scanning electron micrographs of a variety of biofilms are presented in Figures 2 to 6. These figures show the range of catheter biofilms observed, from thin (3/am) layers of cells to thick multi-layers up to 490/am in depth. The critical point-dried preparations all confirm that the biofilms are composed of masses of bacterial cells.
).
Discussion IIHtlOLtl
¢ORU
N100
100~
I
001042
The recent study of Ramsey et al. (6) demonstrated that 16 of 33 Foley catheters that had been indwelling for a modal duration of two to three weeks were colonized by bacterial biofilms. Ohkawa et al. (15) reported that biofilms could be seen on 21 of 57 catheters that had been in place for a range of 1 to 16 days (mean 7.3 days). The greater incidence of catheter colonization observed in this study probably results from the longer duration of catheterization (3 to 83 days, mean 35 days) and the fact that these catheters were taken from groups of patients with established chronic infections undergoing longterm catheterization. There is no simple relationship between depth of biofilm and duration of catheterization. Regression analysis revealed a correlation coefficient (R 2) of 0.019 (not significant, p > 0.05) between biofilm depth and time in situ for the data presented in Figure 1.
Figure 3: (a) Freeze-dried cross-section of a 42-day-old catheter colonized by a mixed population of Escherichia coli (6.5 x 107 cfu/cm2), Klebsiella pneumoniae (4.6 x 106 cfu/cm2). (b) Critical point-dried preparation showing the biofilm surface structure.
Ohkawa et al. (15) observed that biofilms were not visible on catheters that had been indwelling for less than seven days, whereas 13 of the 14 that had been in place for two weeks or more were colonized. It is clear from their data, however, that they were dealing with a population of patients who were not bacteriuric at the time of
Vol. 11, 1992
/
•
793
/
g
e
U
r
e
4: (a) Freeze-dried cross-section of a 42-day-old catheter colonized by Escherichia coil (2.6 x 108 cfu/em2) and bsiella oxytoca (7.9 x 107 cfu/em2). The dotted lines indicate the area shown in higher magnification in (e). (b) Critical point-dried preparation showing the surface of the biofilm. (e) Compilation of micrographs taken at high power across the entire depth of the freeze-dried biofilm shown in (a). The surface of the catheter material (designated CA) is visible in the bottom right hand corner of the figure and biofilm extends as a multi-layered cellular structure some 490 lam into the catheter lumen (designated LU).
794
Eur. J. Clin. Microbiol. Infect. Dis.
,f
't e k o
x
o
~i,B~m
e3~36
a
Figure 6: (a) Freeze-dried cross-section of a 63-day-old catheter containing a mixed population of Staphylococcus epidermMis (3.8 x 104 cfu/cm 2) and Klebsiella oxytoca (1.4 x 104 cfu/cm2). (b) Critical point-dried preparation showing the surface of the biofilm.
Figure 5: (a) Freeze-dried cross-section of a 41-day-old catheter colonized by Morganella morganii (2.4 x 107 cfu/cm 2) and diphtheroids (2.8 x 105 cfu/cm2). (b) Critical point-dried preparation showing the surface of the biofilm. (c) Surface view of the biofilm showing its patchy composition.
catheterization, most of the urine samples taken before day 6 being uninfected. In the patients we have studied established infections are normally present at the time of catheter change. Under these circumstances new biofilms can build up rapidly, for example one catheter that had been in place for just three days was found to be heavily colonized by a film some 50 ~m deep and composed of Escherichia coli, Citrobacter and Pseudomonas aeruginosa. This rapid colonization of silicone surfaces has also been observed with cultures of urinary pathogens growing in urine in models of the catheterized bladder (13). The range of microbial species isolated from the catheters is similar to that reported by other
Vo1.11,1992
workers (6, 7,15). The gram-negative nosocomial species such as Escherichia coli, Klebsiella spp.,
Pseudomonas aeruginosa, Enterobacter cloacae, Proteus and Providencia spp. were rather more prevalent than in the other studies. This reflects the nature of the urinary tract infections characteristic of the groups of patients studied (18). A s has been reported by other groups the microbial flora of the catheter was often different from that of the urine (6, 15, 19). Examination of Table 1 reveals that organisms present in the biofilm were not always recovered from the urine. This was particularly noticeable with Pseudomonas aeruginosa, which was isolated from nine of the catheters but in only four of these cases was it present in the corresponding urine. Ohkawa et al. (15) reported that this species was present on 12 catheters but in only seven cases was it also recovered from the urine. This species is known to be especially adherent to silicone catheters (20) and appears to have a strong preference for growing as a biofilm. The data presented in Table 2 shows that of the 50 patients, 37 had bacteriuria (> 103 cfu/ml) and all of these had catheter biofilms. Of the 13 patients who did not have bacteriuria, nine were undergoing antibiotic therapy at the time of sampling. It ~s particularly interesting that in four of these cases, organisms were surviving well in the catheter biofilm. These observations provide further support for the hypothesis that the catheter biofilm is responsible for the rapid re-emergence of infection that occurs in these patients when antibiotic treatment is completed (10). Ramsay et al. (6) reported that half of the catheters with visible biofilms were colonized by single species, the rest producing two organisms on culture of the catheter. In the present study most of the catheters were colonized by mixed COmmunities of up to four organisms. Communities such as Escherichia coli and Klebsiella oxytoca, Pseudomonas aeruginosa and a group D streptococcus, or Escherichia coli, Citrobacter diversus and a group D streptococcus or even
Enterobacter cloacae, Klebsiella oxytoca, Staphylococcus aureus and a group D streptococcus were isolated. It would be interesting to reconstitute some of these mixed population films in laboratory models and observe their response to the challenge of antibacterial agents. Previous studies using scanning electron microsCOpy on urinary catheter biofilms have provided surface views of the luminal films, from which it has been difficult to interpret their depth. Figures
795
2 to 6, however, give a good insight into the multilayer structure of these films. Figure 2a presents a cross-section of an eight-day catheter. The freezedrying involved in the preparation of these samples has obviously produced some shrinkage of the film but it is clearly visible as a substantial layer (90-180 pm thick) completely coating this region of the catheter lumen. Examination of the surface of this biofilm using critical point-dried preparations (Figure 2b) shows that the film is a mass of bacilli. Culture of the disrupted film revealed a three-membered community of Pseudornonas aeruginosa, Proteus mirabilis and Providencia stuartii. At the time of catheter removal the patient was being treated with gentamicin for a urinary tract infection. Small numbers of Pseudomonas aeruginosa (100/ml) were isolated from the urine but large numbers of each of the four organisms (108/cm 2) were recovered from the catheter. Conventional antibiotic sensitivity tests revealed that all four organisms were sensitive to gentamicin. This is a similar finding to that of Nickel et al. (14) and reinforces the point that gentamicin sensitive organisms can flourish in biofilms in the presence of gentamicin and be a source of reinfection once the antibiotic therapy is completed. It also suggests that when urinary tract infection in these patients is treated, the colonized catheter should be removed and a new one inserted. A catheter that had been in place for 42 days and was colonized by Escherichia coli and Klebsiella pneumoniae is shown in Figure 3. This patient was receiving prophylactic trimethoprim. The thickest film we observed (490 lain) was colonizing a 42-day-old catheter (Figure 4). It was composed of Escherichia coli and Klebsiella oxytoca and both the critical point-dried preparation and the high power micrographs of the interior of the freeze-dried film demonstrate that it is made up of a mass of bacilli packed in a matrix. The high power micrographs (Figure 4c) suggest that the film must be at least 400 cells deep. Some of the catheters were colonized by less substantial biofilms. For example the micrographs presented in Figure 5 are of a 41-day-old catheter from a patient who had significant bacteriuria with Morganella morganii. The biofilm was only a maximum of 10/am thick and the surface views confirm its patchy nature. The other species present in the biofilm was a diphtheroid rod. Figures 6a and b are micrographs of a 63-day-old catheter. The biofilm was composed of Staphylococcus epidermidis and KlebsieUa oxytoca but was
796
only 5 - 1 0 / a m in depth. The surface views also show the sparse nature of this particular film on the catheter lumen. These results and the data summarized from the eight catheters colonized by diphtheroids or coagulase-negative staphylococci (Table 3) suggest that the presence of these organisms might inhibit the development of extensive biofilms by the gram-negative urinary tract pathogens. It would be interesting to test this hypothesis in models of the catheterized bladder. In conclusion, it is clear that when establishing models of bladder catheter biofilms in the laboratory it is important to realize that the natural catheter biofilms can vary from simple patchy monolayers of a single species to extensive mixed community surface coatings up to 490 pm in thickness and some 400 ceils deep.
Acknowledgements We wish to thank Smith Kline and Beecham for their support of this project in the form of a sponsored postgraduate studentship for Liese Ganderton. We are also grateful tO the nursing and medical staffs at the spinal unit of Rookwood Hospital and the urological outpatients clinic at the Cardiff Royal Infirmary who provided the catheters and the information about their history.
References 1. Meers PD, Ayliffe GAJ, Emmerson AM, Leigh DA, Mayon-White RT, Mackintosh CA, Slronge JL: National survey of infection in hospitals 1980. 2: Urinary tract infection. Journal of Hospital Infection 1981, Supplement 2: 23--28. 2. Christensen GD, Purist JT, Bisno AL, Simpson WA, Beachey EH: Characterization of clinically significant strains of coagulase-negative staphylococci. Journal of Clinical Microbiology 1983, 18: 258-259. 3. West TE, Walshe JJ, Krol CP, Amsterdam D: Staphylococcal peritonitis in patients on continuous peritoneal dialysis. Journal of Clinical Microbiology 1986, 23: 809-812. 4. Franson TR, Sheth NK, Rose HD, Sohnle PG: Scanning electron microscopy of bacteria adherent to intravaseular chatheters. Journal of Clinical Microbiology 1984, 20: 500-505. 5. Dasgupta MK, Bettcher KB, Ulan RA, Burns V, Lam K, Dossetor JB, Costerton JW: Relationship of adherent bacterial biofilms to peritonitis in chronic ambulatory peritoneal dialysis. Peritoneal Dialysis Bulletin 1987, 7: 168-173.
Eur. J. Clin. Microbiol. Infect. Dis.
6. Ramsay JWA, Gamham A J, Mulhall AB, Crow RA, Bryan JM, Eardley I, Vale JA, Whitfield HN: Biofitms, bacteria and bladder catheters. British Journal of Urology 1989, 64: 395-398. 7. Nickel JC, Ruseska I, Wright JB, Costerton JW: Tobramycin resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary catheter material. Antimicrobial Agents and Chemotherapy 1985, 27: 619-624. 8. Stickler D J, Claylon CL, Chawla JC: The resistance of urinary tract pathogens to chlorhexidine bladder washouts. Journal of Hospital Infection 1987, 10: 219228. 9. Jensen ET, Kharazmi A, Lain K, Coslerton JW, Hoiby N: Human polymorphonuclear leukocyte response to Pseudomonas aeruginosa grown in biofilms. Infection and Immunity 1990, 58: 2383--2385. 10. Coslerton JW, Cheng KJ, Geesey GG, Ladd TI, Nickel . JC, Dasgupta M, Marrie TJ: Bacterial biofilms in nature and disease. Annual Reviews of Microbiology 1987, 41: 435--464. 11. Stickler D, Dolman J, Rolfe S, Chawla J: Activity of antiseptics against Escherichia colt growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology & Infectious Diseases 1989, 8: 974--978. 12. Stickler D, Dolman J, Rolfe S, Chawla J: Activity of some antiseptics against urinary tract pathogens growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology & Infectious Diseases 1991, 10: 410--415. 13. Stickler D, Hewett P: Activity of antiseptics against biofilms of mixed bacterial species growing on silicone surfaces. European Journal of Clinical Microbiology & Infectious Diseases 1991, 10: 416--421. 14. Nickel JC, Grislina AG, Coslerton JW: Electron microscopic study of an infected Foley catheter. Canadian Journal of Surgery 1985, 28: 50-52. 15. Ohkawa M, Sugata T, Sawaki M, Nakashlma T, Fuse H, Hisazumi H: Bacterial and crystal adherence to the surfaces of indwelling urethral catheters. Journal of Urology 1990, 143: 717-721. 16. Peters AC, Wimpenny JWT: A constant depth laboratory model film fermenter. Biotechnology and bioengineering 1988, 32: 263-270. 17. Cowan ST, Sleel KJ: Manual for the identification of medical bacteria. Cambridge University Press, Cambridge, 1974. 18. Clayton CL, Chawla JC, Stickier DJ: Some observations on urinary tract infections in patients undergoing long-term bladder catheterization. Journal of Hospital Infection 1982, 3: 39--47. 19. Rubin M, Berger SA, Zodda FN, Gruenwald R: Effect of catheter replacement on bacterial counts in urine aspirated from indwelling catheters. Journal of Infectious Diseases 1980, 142: 291. 20. Stickler D J, Clayton CL, Harber MJ, Chawla JC: Pseudomonas aeruginosa and long-term indwelling bladder catheters. Archives o1~ Physical Medicine and Rehabilitation 1988, 69: 25-28.
Vol. 11. No. 9
789
Eur. J. Clin. Microbiol. Infect. Dis., September 1992, p. 789-796 0934-9723/92/09 0789-08 $ 3.00/0
Scanning Electron Microscopy of Bacterial Biofilms on Indwelling Bladder Catheters
L. G a n d e r t o n 1, j. Chawla 2, C. Winters 1, J. W i m p e n n y 1, D. Stickler 1.
Fifty Foley bladder catheters that had been indwelling for periods ranging from 3 to 83 days (mean 35 days) were examined for the presence of bacterial biofilm. Scanning electron microscopy on freeze-dried cross-sections and fixed, critical point-dried longitudinal sections revealed biofilm formation on the luminal surfaces of 44 of the catheters. Culture of urine samples and sonicates from catheters revealed that the prevalence of bacteriuria was less than that of catheter colonization. A wide range of nosocomiai species were found colonizing the catheters, Eschertchia coli being most often isolated. The bacterial composition of the biofilms ranged from single species to mixed communities containing up to four species. There was no relationship between the length of time that the catheter had been in situ and the extent of biofilm formation. The biofilms varied in thickness from 3 to 490/Jm and were visible as layers of bacterial cells up to about 400 cells deep, embedded in a matrix.
Infection is an all too common consequence of intravascular, peritoneal and bladder catheterization (1-3). Examination of catheters taken from these sites has revealed that they are commonly colonized by adherent layers of bacteria embedded in a matrix of extra-cellular polysaccharide (4--6). In this biofilm mode of growth the bacterial cells are more resistant to antibiotics, antiseptics and host defence mechanisms than cells growing in suspension (7-9). These characteristics may well explain why infections associated with these devices are persistent and refractory to treatment with antibacterial agents
(lO).
As part of a programme aimed at understanding the nature and extent of these resistance phenomena we have devised laboratory models of the biofilms that are found on indwelling bladder catheters. Model biofilms have been produced ranging from monolayers of single species to multi-layered films composed of mixed bacterial COmmunities and it has been observed that organisms respond differently to antibacterial agents in films of different thicknesses and composition (11-13). There is a need to obtain more information about the characteristics of the
natural catheter biofilms in order to refine our laboratory models. Previous studies have revealed that urinary catheters can become colonized by single species or mixed populations of organisms. Sparse patchy coverings or confluent thick multi-layered structures have been found in surface views of the catheter lumen (6, 14, 15). One of our laboratory model systems has the facility to produce standard films of defined thickness (16). The objective of the present study was thus to obtain information about the composition and dimensions of the natural biofilms found on indwelling catheters. The previous studies using scanning electron microscopy for examination of urinary catheter biofilms have provided surface views of the luminal films from which it has been difficult to interpret their depth and structure. We have therefore developed a simple freeze-fracturing technique to produce cross-sections Of catheters for scanning electron microscopy.
Materials and Methods Catheters. Silicone or silicone coated Foley catheters
School of Pure and Applied Biology,Universityof Wales 2w~cllegeof Cardiff, PO Box 915, Cardiff CF1 3TL, UK. elsh Spinal Injuries Unit, Rookwood Hospital, Cardiff CF5 2YN, UK,
freshly removed fl'om patients resident in the Welsh Spinal Injuries Unit at Rookwood Hospital and from patients undergoing long-term catheterization and attending a clinic at the Cardiff Royal Infirmary were transported to the laboratory for examination. Details
790
E u r . J. Clin. Microbiol. Infect. Dis.
of the length of time each catheter had been in situ and the nature of any current antibiotic therapy was also obtained together with samples of urine which had been aspirated from the catheters. Viable bacterial cell counts were performed on the urine samples after incubation for 24 h at 37 *C on CLED agar (Oxoid, UK). Characterization o f the Biofihn Communities. To characterize the bacterial communities colonizing the lumen of the catheters, sections (1 cm long) were cut from the region of the catheter within the retention balloon. The sections were rinsed once gently in 10 ml of buffer (Hanks-Hepes, pH 7.4) and then placed in 10 ml of nutrient broth (Oxoid). Disruption of the luminal biofilm was then achieved by sonieation for 5 min (Transsonic Water Bath, Camtab, UK) followed by vortex mixing for 2 min. Samples of the broth suspension were then plated out on CLED agar and incubated for 24 h at 37 *C. The resulting isolates were characterized using Cowan and Steel's tables (17) and the appropriate identification kits
(API, UK.) Scanning Electron Microscopy. Sections of catheters (1 cm in length) were taken from the region adjacent to the retention balloon on the side away from the catheter tip. These were rapidly plunged into liquid nitrogen cooled propane and transferred to liquid nitrogen. Crosssections were then produced by freeze-fracturing samples in a specially designed copper block which held the catheter and a blade in position and facilitated the production of reproducible cross-sections. The samples were then freeze-dried for 24 h at -80 *C, mounted fractured surface uppermost onto aluminium stubs, sputtered with gold and examined in a JEOL JSM5200 scanning electron microscope. To observe the nature of the surfaces of the biofilms longitudinal sections of each catheter were prepared from the region adjacent to the retention balloon. The sections were fixed in 3 % glutaraldehyde in 0.1 M phosphate buffer (pH 7.4) for 1 h and then washed overnight in the phosphate buffer before being post-fixed in Millonig's phosphate buffered osmium tetroxide (1.0 %) for 1 h. The samples were dehydrated in a graded series of
aqueous ethanol solutions (30-100 %). Following dehydration the samples were critical point-dried using liquid CO2. Finally, the samples were mounted on aluminium stubs, sputtered with gold and examined in the scanning electron microscope.
Results Bacterial biofilms were clearly visible on 44 of the catheters. T h e species isolated f r o m t h e 50 paired samples of catheters and urines are listed in Table 1. E s c h e r i c h i a coli was the m o s t c o m m o n biofilm organism, being isolated f r o m 4 0 % of the catheters. Bacteriuria was less c o m m o n than biofilm colonization, a n d m a n y o r g a n i s m s present o n the biofilm w e r e not r e c o v e r e d f r o m the urine. It can be seen f r o m Table 2 t h a t 34 o f the patients w e r e not receiving antibiotics at the time o f c a t h e t e r r e m o v a l and o f this g r o u p 33 had c a t h e t e r biofilms. In five o f the six cases in which no biofilm was r e c o v e r e d the patients w e r e receiving antibiotics at the time of c a t h e t e r r e m o v a l and did n o t h a v e bacteriuria. O f the o t h e r 11 patients u n d e r g o i n g t h e r a p y w h o s e catheters w e r e c o l o n i z e d by biofilms only seven had bacteriuria (> 103 cfu/ml). M o s t o f the biofilms w e r e c o m p o s e d o f mixed bacterial c o m m u n i t i e s , single species being isolated f r o m only 25 % o f the biofilms. T h e p r o p o r tions c o m p o s e d of two, t h r e e a n d f o u r species w e r e 50 %, 20 % and 5 % respectively. Figure 1 presents the d e p t h o f biofilm in relation to the time for which each c a t h e t e r had b e e n in situ. In those cases in which a cross-section o f a
Table 1: Organisms isolated from 50 paired urine and catheter specimens. No. of positive cultures
Organisms
Escherichia coli Group D streptococci Pseudomonas aeruginosa Proteus mirabilis Klebsiella oxytoca Morganella morganii Enterobacter cloacae Staphylococcus epidermidis Providencia spp. Citrobacter diversus Candida spp. Citrobacter freundii Diphtheroids Other species
Catheter
Urine
Both
20 10 9 8 6 6 5 4 4 3 2 2 2 10
17 7 4 6 5 4 6 2 2 3 0 1 2 5
16 7 4 6 4 4 4 2 2 3 0 1 2 5'
Vol. 11, 1992
791
Table 2: Presence of catheter biofilm related to current antibiotic therapy. Antibiotic treatment
Biofilm present, urine positive* Biofilm present, urine negative No biofilm present, urine positive No biofilm present, urine negative
Yes (n = 16)
No (n = 34)
Total (n = S0)
7 4 0 5
30 3 0 1
37 7 0 6
*> 103 cfu/ml.
500,
400.
300.
E 200, .o fl]
•
•
100
•
•e o
•
0 ~ . 0
20
j
4'0
•
6'0
8b"
Time in sifu (days)
Figure 1: Relationship between the thickness of biofilms and the period of time catheters were in situ. catheter revealed a biofilm of variable thickness, the upper limit of the biofilm depth was used in this analysis. In this way Figure 1 reflects the full range of biofilm depths observed. Regression analysis revealed no significant correlation between these parameters. Examination of the extent of biofilm formation in those catheters colonized by coagulase-negative staphylococci and diphtheroid rods (Table 3) shows that despite having been in situ for a mean time of 31 days, these biofilms only had a mean depth of some 15~tm. The equivalent figures observed for catheter biofilms composed of gram-negative bacilli and the enter•cocci were 39 days and 80/am.
Figure 2: (a) Freeze-dried cross-section of an 8-day-old catheter colon~ed by a film consisting of Psettdomonas aerugbtosa (1.1 x 10~ cfu/cm2), Proteus mirabilis (1.2 x 10s cfu/cm2) and Providencia stuartii (5.3 x 107 cfu/emZ). (b) Critical point-dried preparation showing the surface of the biofilm.
792
Eur. J. Clin. Microbiol. Infect. Dis.
Table 3: Characteristics of biofilms containing coagulase negative staphylococci and diphtheroids. Organism
Biofilm depth (gm)
CNS + E. coli Diphtheroids CNS + E. coli CNS + K. oxytoca Diphtheroids + M. morganii CNS CNS CNS + P. rettgeri
3- 5 20-25 5-10 10-15 5-10 20 20-30 5
Days in situ 10 28 12 63 41 42 8 42
CNS = Coagulase-negative staphylococci.
Scanning electron micrographs of a variety of biofilms are presented in Figures 2 to 6. These figures show the range of catheter biofilms observed, from thin (3/am) layers of cells to thick multi-layers up to 490/am in depth. The critical point-dried preparations all confirm that the biofilms are composed of masses of bacterial cells.
).
Discussion IIHtlOLtl
¢ORU
N100
100~
I
001042
The recent study of Ramsey et al. (6) demonstrated that 16 of 33 Foley catheters that had been indwelling for a modal duration of two to three weeks were colonized by bacterial biofilms. Ohkawa et al. (15) reported that biofilms could be seen on 21 of 57 catheters that had been in place for a range of 1 to 16 days (mean 7.3 days). The greater incidence of catheter colonization observed in this study probably results from the longer duration of catheterization (3 to 83 days, mean 35 days) and the fact that these catheters were taken from groups of patients with established chronic infections undergoing longterm catheterization. There is no simple relationship between depth of biofilm and duration of catheterization. Regression analysis revealed a correlation coefficient (R 2) of 0.019 (not significant, p > 0.05) between biofilm depth and time in situ for the data presented in Figure 1.
Figure 3: (a) Freeze-dried cross-section of a 42-day-old catheter colonized by a mixed population of Escherichia coli (6.5 x 107 cfu/cm2), Klebsiella pneumoniae (4.6 x 106 cfu/cm2). (b) Critical point-dried preparation showing the biofilm surface structure.
Ohkawa et al. (15) observed that biofilms were not visible on catheters that had been indwelling for less than seven days, whereas 13 of the 14 that had been in place for two weeks or more were colonized. It is clear from their data, however, that they were dealing with a population of patients who were not bacteriuric at the time of
Vol. 11, 1992
/
•
793
/
g
e
U
r
e
4: (a) Freeze-dried cross-section of a 42-day-old catheter colonized by Escherichia coil (2.6 x 108 cfu/em2) and bsiella oxytoca (7.9 x 107 cfu/em2). The dotted lines indicate the area shown in higher magnification in (e). (b) Critical point-dried preparation showing the surface of the biofilm. (e) Compilation of micrographs taken at high power across the entire depth of the freeze-dried biofilm shown in (a). The surface of the catheter material (designated CA) is visible in the bottom right hand corner of the figure and biofilm extends as a multi-layered cellular structure some 490 lam into the catheter lumen (designated LU).
794
Eur. J. Clin. Microbiol. Infect. Dis.
,f
't e k o
x
o
~i,B~m
e3~36
a
Figure 6: (a) Freeze-dried cross-section of a 63-day-old catheter containing a mixed population of Staphylococcus epidermMis (3.8 x 104 cfu/cm 2) and Klebsiella oxytoca (1.4 x 104 cfu/cm2). (b) Critical point-dried preparation showing the surface of the biofilm.
Figure 5: (a) Freeze-dried cross-section of a 41-day-old catheter colonized by Morganella morganii (2.4 x 107 cfu/cm 2) and diphtheroids (2.8 x 105 cfu/cm2). (b) Critical point-dried preparation showing the surface of the biofilm. (c) Surface view of the biofilm showing its patchy composition.
catheterization, most of the urine samples taken before day 6 being uninfected. In the patients we have studied established infections are normally present at the time of catheter change. Under these circumstances new biofilms can build up rapidly, for example one catheter that had been in place for just three days was found to be heavily colonized by a film some 50 ~m deep and composed of Escherichia coli, Citrobacter and Pseudomonas aeruginosa. This rapid colonization of silicone surfaces has also been observed with cultures of urinary pathogens growing in urine in models of the catheterized bladder (13). The range of microbial species isolated from the catheters is similar to that reported by other
Vo1.11,1992
workers (6, 7,15). The gram-negative nosocomial species such as Escherichia coli, Klebsiella spp.,
Pseudomonas aeruginosa, Enterobacter cloacae, Proteus and Providencia spp. were rather more prevalent than in the other studies. This reflects the nature of the urinary tract infections characteristic of the groups of patients studied (18). A s has been reported by other groups the microbial flora of the catheter was often different from that of the urine (6, 15, 19). Examination of Table 1 reveals that organisms present in the biofilm were not always recovered from the urine. This was particularly noticeable with Pseudomonas aeruginosa, which was isolated from nine of the catheters but in only four of these cases was it present in the corresponding urine. Ohkawa et al. (15) reported that this species was present on 12 catheters but in only seven cases was it also recovered from the urine. This species is known to be especially adherent to silicone catheters (20) and appears to have a strong preference for growing as a biofilm. The data presented in Table 2 shows that of the 50 patients, 37 had bacteriuria (> 103 cfu/ml) and all of these had catheter biofilms. Of the 13 patients who did not have bacteriuria, nine were undergoing antibiotic therapy at the time of sampling. It ~s particularly interesting that in four of these cases, organisms were surviving well in the catheter biofilm. These observations provide further support for the hypothesis that the catheter biofilm is responsible for the rapid re-emergence of infection that occurs in these patients when antibiotic treatment is completed (10). Ramsay et al. (6) reported that half of the catheters with visible biofilms were colonized by single species, the rest producing two organisms on culture of the catheter. In the present study most of the catheters were colonized by mixed COmmunities of up to four organisms. Communities such as Escherichia coli and Klebsiella oxytoca, Pseudomonas aeruginosa and a group D streptococcus, or Escherichia coli, Citrobacter diversus and a group D streptococcus or even
Enterobacter cloacae, Klebsiella oxytoca, Staphylococcus aureus and a group D streptococcus were isolated. It would be interesting to reconstitute some of these mixed population films in laboratory models and observe their response to the challenge of antibacterial agents. Previous studies using scanning electron microsCOpy on urinary catheter biofilms have provided surface views of the luminal films, from which it has been difficult to interpret their depth. Figures
795
2 to 6, however, give a good insight into the multilayer structure of these films. Figure 2a presents a cross-section of an eight-day catheter. The freezedrying involved in the preparation of these samples has obviously produced some shrinkage of the film but it is clearly visible as a substantial layer (90-180 pm thick) completely coating this region of the catheter lumen. Examination of the surface of this biofilm using critical point-dried preparations (Figure 2b) shows that the film is a mass of bacilli. Culture of the disrupted film revealed a three-membered community of Pseudornonas aeruginosa, Proteus mirabilis and Providencia stuartii. At the time of catheter removal the patient was being treated with gentamicin for a urinary tract infection. Small numbers of Pseudomonas aeruginosa (100/ml) were isolated from the urine but large numbers of each of the four organisms (108/cm 2) were recovered from the catheter. Conventional antibiotic sensitivity tests revealed that all four organisms were sensitive to gentamicin. This is a similar finding to that of Nickel et al. (14) and reinforces the point that gentamicin sensitive organisms can flourish in biofilms in the presence of gentamicin and be a source of reinfection once the antibiotic therapy is completed. It also suggests that when urinary tract infection in these patients is treated, the colonized catheter should be removed and a new one inserted. A catheter that had been in place for 42 days and was colonized by Escherichia coli and Klebsiella pneumoniae is shown in Figure 3. This patient was receiving prophylactic trimethoprim. The thickest film we observed (490 lain) was colonizing a 42-day-old catheter (Figure 4). It was composed of Escherichia coli and Klebsiella oxytoca and both the critical point-dried preparation and the high power micrographs of the interior of the freeze-dried film demonstrate that it is made up of a mass of bacilli packed in a matrix. The high power micrographs (Figure 4c) suggest that the film must be at least 400 cells deep. Some of the catheters were colonized by less substantial biofilms. For example the micrographs presented in Figure 5 are of a 41-day-old catheter from a patient who had significant bacteriuria with Morganella morganii. The biofilm was only a maximum of 10/am thick and the surface views confirm its patchy nature. The other species present in the biofilm was a diphtheroid rod. Figures 6a and b are micrographs of a 63-day-old catheter. The biofilm was composed of Staphylococcus epidermidis and KlebsieUa oxytoca but was
796
only 5 - 1 0 / a m in depth. The surface views also show the sparse nature of this particular film on the catheter lumen. These results and the data summarized from the eight catheters colonized by diphtheroids or coagulase-negative staphylococci (Table 3) suggest that the presence of these organisms might inhibit the development of extensive biofilms by the gram-negative urinary tract pathogens. It would be interesting to test this hypothesis in models of the catheterized bladder. In conclusion, it is clear that when establishing models of bladder catheter biofilms in the laboratory it is important to realize that the natural catheter biofilms can vary from simple patchy monolayers of a single species to extensive mixed community surface coatings up to 490 pm in thickness and some 400 ceils deep.
Acknowledgements We wish to thank Smith Kline and Beecham for their support of this project in the form of a sponsored postgraduate studentship for Liese Ganderton. We are also grateful tO the nursing and medical staffs at the spinal unit of Rookwood Hospital and the urological outpatients clinic at the Cardiff Royal Infirmary who provided the catheters and the information about their history.
References 1. Meers PD, Ayliffe GAJ, Emmerson AM, Leigh DA, Mayon-White RT, Mackintosh CA, Slronge JL: National survey of infection in hospitals 1980. 2: Urinary tract infection. Journal of Hospital Infection 1981, Supplement 2: 23--28. 2. Christensen GD, Purist JT, Bisno AL, Simpson WA, Beachey EH: Characterization of clinically significant strains of coagulase-negative staphylococci. Journal of Clinical Microbiology 1983, 18: 258-259. 3. West TE, Walshe JJ, Krol CP, Amsterdam D: Staphylococcal peritonitis in patients on continuous peritoneal dialysis. Journal of Clinical Microbiology 1986, 23: 809-812. 4. Franson TR, Sheth NK, Rose HD, Sohnle PG: Scanning electron microscopy of bacteria adherent to intravaseular chatheters. Journal of Clinical Microbiology 1984, 20: 500-505. 5. Dasgupta MK, Bettcher KB, Ulan RA, Burns V, Lam K, Dossetor JB, Costerton JW: Relationship of adherent bacterial biofilms to peritonitis in chronic ambulatory peritoneal dialysis. Peritoneal Dialysis Bulletin 1987, 7: 168-173.
Eur. J. Clin. Microbiol. Infect. Dis.
6. Ramsay JWA, Gamham A J, Mulhall AB, Crow RA, Bryan JM, Eardley I, Vale JA, Whitfield HN: Biofitms, bacteria and bladder catheters. British Journal of Urology 1989, 64: 395-398. 7. Nickel JC, Ruseska I, Wright JB, Costerton JW: Tobramycin resistance of Pseudomonas aeruginosa cells growing as a biofilm on urinary catheter material. Antimicrobial Agents and Chemotherapy 1985, 27: 619-624. 8. Stickler D J, Claylon CL, Chawla JC: The resistance of urinary tract pathogens to chlorhexidine bladder washouts. Journal of Hospital Infection 1987, 10: 219228. 9. Jensen ET, Kharazmi A, Lain K, Coslerton JW, Hoiby N: Human polymorphonuclear leukocyte response to Pseudomonas aeruginosa grown in biofilms. Infection and Immunity 1990, 58: 2383--2385. 10. Coslerton JW, Cheng KJ, Geesey GG, Ladd TI, Nickel . JC, Dasgupta M, Marrie TJ: Bacterial biofilms in nature and disease. Annual Reviews of Microbiology 1987, 41: 435--464. 11. Stickler D, Dolman J, Rolfe S, Chawla J: Activity of antiseptics against Escherichia colt growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology & Infectious Diseases 1989, 8: 974--978. 12. Stickler D, Dolman J, Rolfe S, Chawla J: Activity of some antiseptics against urinary tract pathogens growing as biofilms on silicone surfaces. European Journal of Clinical Microbiology & Infectious Diseases 1991, 10: 410--415. 13. Stickler D, Hewett P: Activity of antiseptics against biofilms of mixed bacterial species growing on silicone surfaces. European Journal of Clinical Microbiology & Infectious Diseases 1991, 10: 416--421. 14. Nickel JC, Grislina AG, Coslerton JW: Electron microscopic study of an infected Foley catheter. Canadian Journal of Surgery 1985, 28: 50-52. 15. Ohkawa M, Sugata T, Sawaki M, Nakashlma T, Fuse H, Hisazumi H: Bacterial and crystal adherence to the surfaces of indwelling urethral catheters. Journal of Urology 1990, 143: 717-721. 16. Peters AC, Wimpenny JWT: A constant depth laboratory model film fermenter. Biotechnology and bioengineering 1988, 32: 263-270. 17. Cowan ST, Sleel KJ: Manual for the identification of medical bacteria. Cambridge University Press, Cambridge, 1974. 18. Clayton CL, Chawla JC, Stickier DJ: Some observations on urinary tract infections in patients undergoing long-term bladder catheterization. Journal of Hospital Infection 1982, 3: 39--47. 19. Rubin M, Berger SA, Zodda FN, Gruenwald R: Effect of catheter replacement on bacterial counts in urine aspirated from indwelling catheters. Journal of Infectious Diseases 1980, 142: 291. 20. Stickler D J, Clayton CL, Harber MJ, Chawla JC: Pseudomonas aeruginosa and long-term indwelling bladder catheters. Archives o1~ Physical Medicine and Rehabilitation 1988, 69: 25-28.
