High Resolution Imaging Reveals Microbial Biofilms on Patient Urinary Catheters Despite Antibiotic Administration

Abstract

Purpose:

Catheter-associated urinary tract infections (CAUTIs) are a significant cause of morbidity worldwide, as they account for 40% of all hospital-associated infections. Microbial biofilm formation on urinary catheters (UCs) limits antibiotic efficacy, making CAUTI extremely difficult to treat. To gain insight into the spatiotemporal microbe interactions on the catheter surface we sought to determine how the presence or absence of bacteriuria prior to catheterization affects the organism that ultimately forms a biofilm on the UC and how long after catheterization they emerge.

Methods:

Thirty UCs were collected from patients who received a urine culture prior to catheterization, a UC, and antibiotics as part of standard of care. Immunofluorescence imaging and scanning electron microscopy were used to visualize patient UCs.

Results:

Most patients did not have bacteria in their urine (based on standard urinalysis) prior to catheterization, yet microbes were detected on the majority of UCs, even with dwell times of <3 days. The most frequently identified microbes were Staphylococcus epidermidis, Enterococcus faecalis, and Escherichia coli.

Conclusions:

This study indicates that despite patients having negative urine cultures and receiving antibiotics prior to catheter placement, microbes, including uropathogens associated with causing CAUTI, could be readily detected on UCs with short dwell times. This suggests that a potential microbial catheter reservoir can form soon after placement, even in the presence of antibiotics, which may serve to facilitate the development of CAUTI. Thus, removing and/or replacing UCs as soon as possible is of critical importance to reduce the risk of developing CAUTI.

INTRODUCTION

Urinary catheters are the most commonly placed medical devices in the US, with as many as ~16% of hospitalized patients and 5% of long-term care facilities residents receiving an indwelling catheter [1, 2]. Urinary catheterization strikingly increases the risk of developing asymptomatic catheter-associated bacteriuria (ACAB) and symptomatic catheter-associated urinary tract infections (CAUTIs), especially to atypical uropathogens, such as Enterococcus spp., Pseudomonas spp., and Staphylococcus spp [2–5]. While ACAB is not considered a clinically significant condition and antibiotics should not be administered [2–4, 6], CAUTIs are a significant healthcare concern. They escalate healthcare costs, extend hospital stays, and contribute to increased patient morbidity and mortality, as they can result in severe complications, including pyelonephritis, bacteremia, and urosepsis [2, 4–8]. Thus, strategies that prevent and/or treat CAUTIs are essential to limit the most serious complications that occur [2–4, 6]. Recent studies suggest that stringent guidelines that limit unnecessary catheter use (catheters placed without need or indication) are effective at reducing CAUTI rates [9–11]. However, for patients that require a catheter, CAUTI rates are high [2]. Problematically, biofilm formation on catheters makes treatment difficult, as these communities provide a physical barrier through the incorporation of both host and microbial factors, which reduce antibiotic efficacy [4, 5, 12]. Furthermore, guidelines to differentiate ACAB from CAUTIs are complex, as they require individuals to manifest a combination of symptoms within two days of catheter placement or removal, including but not limited to: fever, suprapubic or costovertebral angle pain, urinary urgency or frequency, and dysuria [13, 14]. The manifestation of these symptoms can very widely among different people, making it difficult to determine which individuals have ACAB and which should be treated for CAUTI [2, 15–17]. Thus, to develop effective prevention and/or treatment strategies, a better understanding of the host-pathogen interactions that lead to CAUTI are needed.

In this study we used standard microbiological culturing techniques, immunofluorescence imaging, and microscopic techniques to: i) establish whether microbes causing bacteriuria prior to catheterization also colonize the patient’s catheter following placement; ii) identify microbes present on urinary catheters from patients that did not have bacteriuria prior to catheterization; and iii) determine the effect of antibiotic administration on microbial biofilm formation on catheters, to better understand the spatiotemporal catheter-microbe interactions.

METHODS

Urinary catheter collection.

Informed consent was obtained and patient catheters were collected once the clinical decision to remove for standard of care was made, as previously described [18]. A patient population that received a urine culture prior to catheterization was targeted, which allowed us to differentiate patients with bacteriuria from those that did not have detectable levels of bacteria in their urine prior to catheterization. The Barnes Jewish Hospital Clinical Microbiology Lab (BJH) performed all urine culture analyses. Only pre-catheter urine was assessed. Of note, these patients received prophylactic antibiotic therapy per standard of care recommendations, as the majority underwent urologic interventions [3]. This study was approved by the Washington University School of Medicine (WUSM) Internal Review Board (approval #201410058) and performed in accordance with WUSM’s ethical standards and the 1964 Helsinki declaration and its later amendments.

Microbial identification and antibiotic susceptibility testing.

The catheter tip (~0.5 cm) was cultured in brain heart infusion (BHI) broth (24–48 hrs, shaking at 37°C) and then restreaked onto a BHI agar plate for single colonies. For identification, a single colony of each type based on morphology, size, and color was selected, DNA was extracted, and 16S sequencing was performed using the primers listed in Supplementary Table S1, as previously described [19]. Minimum inhibitory concentrations (MICs) were assessed as previously described [20] for uropathogens isolated from patient catheters (see the supplemental materials (SM) for more details).

Immunofluorescence analysis (IFA).

The remaining 10 cm of the catheter tip was fixed in 10% neutral buffered formalin, with the first 3 cm processed for scanning electron microscopy (SEM) analysis (see SEM analysis) and the next 7 cm assessed by IFA, as previously described [18, 19, 21]. Briefly, for IFA, catheters were blocked, stained with primary and secondary antibodies, and imaged using the Odyssey Imaging System (LI-COR Biosciences). Negative controls included catheter pieces treated identically, with the exception that primary antibodies were excluded (more details in SM).

SEM analysis.

Catheter tips were processed for SEM as previously described [19]. Briefly, catheters were washed in 0.15 M sodium cacodylate buffer, stained with 1% osmium tetroxide, and dehydrated in a graded ethanol series (50, 70, 90, 100, 100%). Samples were critical-point dried (Leica CPD 300; Vienna, Austria), sputter coated with 6 nm iridium (Leica ACE 600; Vienna, Austria), and imaged on a Zeiss Merlin FE-SEM equipped with a Gemini II electron column (Oberkochen, Germany), with secondary electrons (SE) detected by the Everhart-Thornley SE2 detector (accelerating voltage: 3 keV, beam current: 200 nA) (more details in SM).

RESULTS

Microbes Cultured from Patient Urinary Catheters.

Thirty adult patients, representing both males and females, undergoing urinary catheterization for standard of care in the Division of Urology at WUSM were consented for this study once the clinical decision to remove the catheter was made. The patient gender, catheter dwell time, urine culture results, reason for catheterization, and antibiotics prescribed are shown in Table 1. Catheters had a range of dwell times, from <1 to ~28 days.

Table 1.

Data recorded for urinary catheters placed as standard care.

Pt #	Gender	Dwell time (days)	Urine culture	Procedure	Antibiotics prescribed
1	F	1	+	PCNL	Ciprofloxacin BID 7 days prior to placement IV Cefazolin & Ciprofloxacin IO and PO
2	M	1	-	Partial NERF	IV Cefazolin IO
3	M	8	-	RRP	IV Cefazolin & gentamicin IO; IV Cefazolin PO
4	M	8	+	Partial Penectomy	IV Cefepime IO
5	M	9	-	Urethroplasty	IV Gentamicin & Cefazolin IO; IV Cefazolin PO
6	M	15	+	Urethroplasty	IV Gentamicin & Cefazolin IO; IV Cefazolin & Ciprofloxacin PO Bactrim BID for 5 days
7	M	28	-	RRP	IV Cefazolin IO; Ciprofloxacin PO Bactrim BID
8	M	4	-	TURBT	IV Ciprofloxacin IO
9	M	27	-	RRP	IV Cefazolin IO; Ciprofloxacin PO
10	F	1	-	NEFR	IV Cefazolin IO
11	M	1	-	Artificial Sphincter	IV Vancomycin & Ampicilin/Sulbactam IO
12	M	1	+	TURBT	IV Ciprofloxacin IO; Ciprofloxacin PO
13	M	1	-	NEFR	IV Cefazolin IO
15	M	6	-	Lung biopsy	IV Cefazolin PreOp IV Cefazolin & Ciprofloxacin BID PO Ciprofloxacin PO BID for 7 days
17	M	3	-	Neurogenic bladder	IV Cefriaxone IO; then IV Meropenem; Ciprofloxacin PO BID for 6 days.
18	M	11	-	RRP	IV Cefazolin IO; Ciprofloxacin OP
19	M	8	+	RRP	IV Cefazolin IO; Levaquin PO
20	M	2	-	Groin abscess drainage and debridement	IV Vancomycin, Clindamycin & Aztreonam IO; IV Clindamycin, Vancomycin, & Meropenem PO
21	M	15	-	RRP	IV Vancomycin & Gentamicin IO; Ciprofloxacin PO
22	M	7	-	RRP	IV Cefazolin IO; Ciprofloxacin PO
23	M	7	-	RRP	IV Cefazolin IO; Ciprofloxacin PO
24	M	2	-	LRN	Neomycin & Metronidazole BID PO IV Cefazolin & Ertapenem PO
25	M	2	-	Urinary retention	IV Ciprofloxacin & Ceftriaxone IO
26	M	2	-	RAPN	IV cefazolin IO and PO
27	M	1	-	RAPN	IV Cefazolin PreOp and PO
28	M	1	-	Artificial sphincter	IV Vancomycin & Ampicillin/Sulbactam PO
29	M	1	-	Replacement of Artificial sphincter	IV Vancomycin & Ampicillin/Sulbactam IO and PO
30	F	3	-	Partial NEFR	IV Cefazolin IO and OP

F: female

M: male

PCNL: percutaneous nephrolithotomy

NERF: nephrectomy

LRN: laparoscopic radical NERF

RRP: radical retropubic prostatectomy

TURBT: transurethral resection of bladder tumor

RAPN: robotic-assisted partial nephrectomy

BID: on prescription

IV: intravenous

IO: intraoperatively

OP: postoperatively

Patients’ pre-catheterization urine and catheter culture results are shown in Table 2. Only 5/30 patients had positive urine cultures. Of those, one patient had a subsequent negative catheter culture, three had successive catheter cultures concomitant for the same uropathogen found in their pre-catheterization urine culture, which included Escherichia coli, Pseudomonas aeruginosa, and “Gram positive bacteria”, and one had a discordant catheter culture positive for Candida spp. However, the vast majority of patients (25/30) had negative pre-catheterization urine cultures (no detectable bacteria in their urine based on standard urinalysis). Nine patients also had concomitant negative catheter cultures (Supplementary Figure S1). However, both typical and atypical uropathogens were cultured from 64% (16/25) of catheters from patients with negative urine cultures. Staphylococcus epidermidis was the most commonly isolated species (5 catheters), followed by E. coli and Enterococcus faecalis (4 catheters each), P. aeruginosa (3 catheters), Staphylococcus aureus, Corynebacterium spp., and Candida spp. (2 catheters each), and Proteus spp., Micrococcus spp, Macrococcus spp., Lactococcus rhamnosus, and Staphylococcus haemolyticus (1 catheter each). To determine antibiotic susceptibility of the uropathogens isolated from each patient’s catheter, MIC assays were performed. The uropathogens tested were sensitive to at least one antibiotic administered to the respective patient (Supplementary Table S2).

Table 2.

Comparison of Microbes detected from urine and catheters cultures.

Pt #	Dwell time (days)	Pre-catheter urine culture	Organism	Catheter culture	Organism
1	1	+	E. coli (>50,000 CFU/ml)	+	E. coli
2	1	-	n/a	-	n/a
3	8	-	n/a	+	E. faecalis
4	8	+	P. aeruginosa (>50,000 CFU/ml)	+	P. aeruginosa
5	9	-		+	E. faecalis S. epidermidis
6	15	+	Coagulase negative staphylococcus (>5,000 colonies/ml)	-	n/a
7	28	-	n/a	+	P. aeruginosa S. aureus
8	4	-	n/a	+	S. epidermidis
9	27	-	n/a	+	E. coli
10	1	-	n/a	+	E. faecalis
11	1	-	n/a	-	n/a
12	1	+	K. oxytoca (>50,000 CFU/ml) E. faecalis (>5,000 CFU/ml)	+	Candida spp.
13	1	-	n/a	+	S. haemolyticus
14	1	-	n/a	+	S. aureus
15	6	-	n/a	+	Candida spp.
16	9	-	n/a	+	Proteus spp. S. epidermidis
17	3	-	n/a	+	S. epidermidis
18	11	-	n/a	-	n/a
19	8	+	Gram + organism (<5,000 CFU/ml)	+	S. epidermidis Corynebacterium
20	2	-	n/a	+	S. epidermidis
21	15	-	n/a	-	n/a
22	7	-	n/a	+	E. coli
23	7	-	n/a	+	E. coli P. aeruginosa
24	2	-	n/a	-	n/a
25	2	-	n/a	+	Corynebacterium
26	2	-	n/a	+	Micrococcus spp. Macrococcus spp.
27	1	-	n/a	-	n/a
28	1	-	n/a	-	n/a
29	1	-	n/a	-	n/a
30	3	-	n/a	+	E. faecalis L. rhamnosus

Escherichia coli, Pseudomonas aeruginosa, Klebsiella oxytoca, and Enterococcus faecalis were isolated from urine cultures.

Escherichia coli, Enterococcus faecalis, Pseudomonas aeruginosa, Staphylococcus epidermidis, Staphylococcus aureus, Candida, Proteus, Corynebacterium, Staphylococcus haemolyticus, Micrococcus, Macrococcus, and Lactobacillus rhamnosus were isolated from catheter cultures.

Microbes Detected on Patient Urinary Catheters.

Microbes were detected on patient urinary catheters via IFA, using commercially available antibodies (antibodies were not available for Corynebacterium, Lactococcus, or Micrococcus/Macrococcus spp.). The microbes identified in the urine (5) or catheter (19) cultures were present on all patient catheters (Figure 1). The majority of catheters (13) had a single microorganism present, including E. coli (catheters 1, 9, and 22), E. faecalis (catheters 3 and 10), P. aeruginosa (catheter 4), S. epidermidis (catheter 8, 17, and 20), Candida spp. (catheter 12 and 15), S. haemolyticus (catheter 13), and S. aureus (catheter 14). Four catheters had more than one microorganism detected, including E. faecalis and S. epidermidis (catheter 5), P. aeruginosa and S. aureus (catheter 7), Proteus spp. and S. epidermidis (catheter 16), and E. coli and P. aeruginosa (catheter 23).

Figure 1.

Images of Immunofluorescencely stained patient urinary catheters that had positive pre-catheter urine or post-catheterization catheter cultures. Patient catheters were stained with commercially available antibodies for the respective microorganism detected by standard microbiologic techniques. Catheter analysis revealed that Escherichia coli (EC), Enterococcus faecalis (EF), Pseudomonas aeruginosa (PS), Staphylococcus epidermidis (SE), Staphylococcus aureus (SA), Candida albicans (CA), Staphylococcus hemolyticus (SH), and Proteus mirabilis (PT) were detected on catheters, which corresponded to their respective cultures. Human catheters without primary antibody served as controls.

Visualizing Microbes on Patient Urinary Catheters.

Patients with positive cultures (22) were selected for SEM analysis to better visualize the microbial-catheter interactions (Figure 2A). Catheter 4 had biofilm structures consistent with P. aeruginosa, which was cultured from the urine and catheter. Despite having a negative catheter culture, individual cocci were present on catheter 6, which was consistent with the “coagulase negative staphylococci” found in the urine. This suggests that the prescribed antibiotics killed these bacteria or that they were in a non-culturable state. Additionally, biofilm structures consistent with P. aeruginosa and S. aureus (catheter 7) and pseudo-hyphae consistent with Candida spp. (catheter 15) were visible, which correlated with the respective catheter cultures. Cocci consistent with S. epidermidis (catheters 19 and 20) and Macrococcus spp. (catheter 26) and rods matching Corynebacterium spp. (catheter 25) and L. rhamnosus (catheter 30), were present. Of note, microbes could only be visualized via SEM on nine of 22 catheters. One limitation may have been that biofilm was not present on the small catheter piece analyzed (~3 cm), as Figure 1 indicates not all catheter areas contained microbes. Furthermore, the extensive deposition of host factors, including immune cells, epithelial cells, and other host debris, on catheters made visualization of microbes difficult (Figure 2B). Importantly, this supports other studies that show the catheter surface is altered following placement [2–4, 18, 19, 21].

Figure 2.

Representative scanning electron images of human urinary catheters. A) Black arrowheads denote Pseudomonas aeruginosa (PS) on catheter #4, coagulase-negative staphylococci (CNS) on catheter #6, Staphylococcus aureus (SA) on catheter #7, Candida spp. (Can) on catheter #15, Staphylococcus epidermidis (SE) on catheter #19, S. epidermidis (SE) on catheter #20, Corynebacterium spp. (Coryn) on catheter #25, Macrococcus spp. (Mac) on catheter #26, and Lactobacillus rhamnosus (LR) on catheter #30. Asteric denotes P. aeruginosa (PS) on catheter #7. Scale bars are 2 um for all catheters expect catheter #15, which is 10 um. B) Images show representative catheters # 1, # 3, and #16 with host factors, including host proteins (<), immune cells (‣), epithelial cells (*), and other host debris (+). Scale bar is 10 um.

DISCUSSION

This study shows that most patients had negative pre-catheterization urine cultures; yet microbes, including uropathogens, were present on the majority of collected catheters, even after short dwell times (<3 days). The most frequently detected bacterium was S. epidermidis, which rarely causes CAUTI [2]. Additionally, Lactobacillus and Corynebacterium spp., which are members of the predicted urinary microbiome [22], were detected on three catheters. This supports other studies that show catheter colonization is not always problematic and does not always result in symptomatic infection [3, 23]. It also suggests that the urinary microflora infrequently colonize urinary catheters. Most patient urinary catheters were colonized with typical CAUTI uropathogens, including Proteus spp, E. coli, E. faecalis, and P. aeruginosa [2]. Additionally, this study demonstrates that microbial biofilm forms on patient catheters, despite antibiotic administration prior to catheterization, suggesting that i) antibiotics are ineffective at preventing catheter colonization and ii) a potential catheter reservoir of uropathogens develops soon after placement in most patients who receive an indwelling catheter. Thus, this data may explain why recent studies show that even empiric treatment of either ACAB and/or CAUTI does not improve patient outcomes, as once they manifest, these microbes have already formed recalcitrant biofilms that make antibiotics less effective [2–6, 18, 21]. Lastly, the role of prophylactic antibiotic treatment warrants careful study, as it is becoming clear that subinhibitory concentrations can stimulate biofilm formation in many bacterial species, including E. coli and P. aeruginosa, and can make subsequent infections more difficult to treat [24]. Thus, these data strongly reinforce that antibiotics are not an effective intervention at preventing catheter colonization, and should be highly discouraged to prevent the development of antibiotic resistant microbes. It also highlights the importance of removing and replacing urinary catheters as soon as possible or implementing clean intermittent catheterization protocols over chronic indwelling catheters, when possible, to eliminate the reservoir of uropathogens, which would likely i) reduce the risk of ACAB progressing to CAUTIs and ii) increase antibiotic efficacy once CAUTIs manifest.

The lack of effective treatment options and the inability to identify which patients will go on to develop CAUTIs or other complications remain a critical public health problem, as CAUTIs can result in severe morbidity and mortality [2, 4–6]. The current work provides insights into the identity of and the timing of the primary microbes colonizing patient catheters; however, future studies with a larger cohort that includes individuals outside the urologic population are needed to determine the generalizability of these results to females and outpatients. Additionally, this work may explain why antimicrobial coatings are not effective at preventing CAUTI with chronic indwelling catheters [2], as the progressive deposition of host and bacterial factors on catheter surfaces may provide a physical barrier that shields the microbes.

CONCLUSIONS

This work begins to dissect the spatiotemporal catheter-microbe interactions that may lead to CAUTI development. We found that microbes colonize catheters soon after placement, despite patients having negative urine cultures prior to catheterization and receiving prophylactic antibiotics. This work also highlights the fact that catheter colonization is likely only one step in the progression to symptomatic infection and other determinants, such as the host response or bacterial virulence, are equally as important. Future studies that capture symptomatic events in catheterized patients will be essential for elucidating the host-pathogen mechanisms that facilitate persistent, symptomatic infections. These molecular studies are desperately needed to provide the insights required to develop more effective prevention and therapeutic strategies that improve outcomes in catheterized patients.