Bacterial Interference for Prevention of Urinary Tract Infection

Transcription

1 HEALTHCARE EPIDEMIOLOGY INVITED ARTICLE Robert Weinstein, Section Editor Bacterial Interference for Prevention of Urinary Tract Infection Rabih O. Darouiche 1,2,3,4 and Richard A. Hull 5 1 Medical and Spinal Cord Injury Care Lines, Infectious Disease Section, Michael E. DeBakey Veterans Affairs Medical Center, and Departments of 2 Medicine, 3 Surgery, and 4 Physical Medicine and Rehabilitation, and 5 Department of Molecular Virology and Microbiology, Baylor College of Medicine, Houston, Texas The concept of bacterial interference, both passive and active, embodies the use of bacteria of low virulence to compete with and protect against colonization and infection by disease-causing organisms. Passive bacterial interference is achieved when naturally present commensal bacteria help defend against host infection by pathogenic organisms. Examples of bodily sites that are protected by normal flora include the skin, oral cavity, upper respiratory tract, vagina, bladder, and bowels. In instances where natural bacterial flora may not afford adequate resistance to colonization and/or infection, active bacterial interference can be useful by intentionally administering into the human body bacteria of low virulence that could replace pathogens. The potential benefit of bacterial interference in preventing infection has been explored with variable success in differently designed studies of various conditions (including urinary tract infection, vaginosis, burn, and antibiotic-associated colitis) by using different agents (including Escherichia coli, Lactobacillus species, Streptomyces species, and skin flora) [1 3]. Not only is urinary tract infection (UTI) the most common healthcare-acquired infection, but it also accounts for 8 11 million patient visits each year in the United States [4]. Because the risk-benefit analysis does not favor the general use of antimicrobial-based measures Received 30 April 2012; accepted 18 July 2012; electronically published 24 July Correspondence: Rabih O. Darouiche, MD, Baylor College of Medicine, 1333 Moursund Ave, Ste A221, Houston, TX (rdarouiche@aol.com). Clinical Infectious Diseases 2012;55(10): Published by Oxford University Press on behalf of the Infectious Diseases Society of America DOI: /cid/cis639 to prevent UTI, there exists a pressing need to explore the role of nonantimicrobial preventive approaches. To eliminate uropathogens, a number of studies have aimed at deliberate colonization of the bladder with bacteria of low virulence. This approach is supported by the fact that treatment of asymptomatic bacteriuria is not only contraindicated in general, but can also lead to emergence of uropathogens and development of symptomatic infection [5 7]. This review focuses on the role of bacterial interference in preventing UTI by using E. coli in patients with neurogenic/dysfunctional bladder or Lactobacillus species mostly in healthy women. It assesses the role of different agents of bacterial interference, explores the potential mechanisms of action, summarizes the results of clinical studies, compares different bacterial applications, and highlights future research. AGENTS OF BACTERIAL INTERFERENCE The 2 bacterial organisms that have been clinically tested as agents of bacterial interference for specifically preventing UTI include E. coli (E. coli strains and HU2117) and a variety of Lactobacillus species (mostly L. crispatus CTV05, L. casei var rhamnosus GR1 and GG, and L. fermentum B54). Escherichia coli Bacterial interference utilizing nonpathogenic E. coli for prevention of UTI was originally established by using E. coli 83972, an organism that had caused asymptomatic bacteriuria for 3 years in a young girl without altering renal function [8]. Although E. coli has not been reported to cause symptomatic UTI and usually does not express P fimbriae, it contains the full chromosomal pap gene that encodes P fimbriae [9, 10]. Escherichia coli HU2117 is a 1400 CID 2012:55 (15 November) HEALTHCARE EPIDEMIOLOGY

2 genetically engineered strain created from wild-type E. coli that has an 800-base pair deletion mutation in the chromosomal papg gene [9]. As a result, it cannot make functional P fimbriae, which are associated with pyelonephritis and bacteremia. Based on in vitro data, E. coli HU2117 seems to be indistinguishable from E. coli with respect to bladder colonization and ability to prevent biofilm formation on urinary catheters by uropathogens [11]. Lactobacillus species Lactobacilli are the dominant vaginal flora and possess antimicrobial properties that regulate other urogenital microbes. Although a variety of Lactobacillus species have been reported as healthy vaginal flora, L. crispatus, L. gasseri, L. inners, andl. jensenii appear to be most common [12]. Lactobacillus casei var rhamnosus strains have been evaluated as potential probiotic agents in the vagina but have not been reported as part of normal vaginal flora [12]. Since high concentrations of Lactobacillus species in the vagina correlate with a lower rate of UTI, the use of Lactobacillus-containing probiotics to restore commensal vaginal flora has been proposed for the treatment and prophylaxis of bacterial urogenital infections [4, 13]. In most occasions, however, Lactobacillus species is intentionally introduced into the human body for reasons other than prevention of UTI, including of bacterial vaginosis and regulation of bowel flora. Because the potential for preventing UTI using bacterial interference by probiotic Lactobacillus is based primarily on the displacement of potential uropathogens from the vagina, this approach is utilized exclusively in women. An attempt to colonize the urethra or bladder directly through intravesicle inoculation with Lactobacillus was unsuccessful in men [13]. MECHANISMS OF ACTION Figure 1 displays a humorous scientific version of potential mechanisms of bacterial interference. Mechanisms of action, listed in a clockwise fashion starting at the right upper corner, include competition for nutrients, production of antibacterial agents (including bacteriocins, antiseptics, and acidic substances), competition for attachment sites, regulation of gene expression, immunomodulation, and biofilm disruption. Coaggregation of organisms could also enhance the goal of bacterial interference. Escherichia coli and Lactobacillus species share some but not all mechanisms of bacterial interference. Escherichia coli Competition for Nutrients It has been suggested that the primary mechanism for bacterial interference by E. coli is competition with other bacteria for nutrients [14]. This concept was based on the observation that E. coli grew faster than several tested uropathogenic E. coli strains and significantly outperformed 1 uropathogenic E. coli strain in cocultures, both in vitro and in a murine bladder colonization model. However, human colonization trials indicate that E. coli may not eliminate other bacteria from the urine after relatively few generations. Figure 1. Potential mechanisms of bacterial interference: The Bladder Inn. Abbreviations: G, good; B, bad. HEALTHCARE EPIDEMIOLOGY CID 2012:55 (15 November) 1401

3 During long-term colonization in human subjects, E. coli colonized the bladder along with other bacteria that coexisted at the same or lower concentration [15 17]. In some instances bladder cocolonization along with E. coli was transient, whereas in others cocolonizers persisted for extended periods of time in the absence of symptomatic UTI. Bacteriocin Production Bacteriocins, including colicins and microcins, are antibiotic proteins produced by many species of bacteria that kill susceptible bacteria of the same species. An in vitro study showed that coating a urinary catheter with a colicin-producing strain of E. coli completely prevented catheter colonization by a susceptible clinical isolate but not by a colicin-resistant isolate [18]. Escherichia coli does not produce detectable levels of bacteriocin when studied in vitro, but genomic studies suggest that it has the genetic potential for expressing 3 microcins, namely, H47, M, and B17 [19]. Competition for Attachment Sites A recent report suggested that attachment of probiotic E. coli Nissle to host tissue may prevent attachment of pathogenic bacteria [20]. Although E. coli Nissle has been used for prevention of various intestinal diseases, its use as a probiotic agent for prevention of UTI has been proposed but not evaluated experimentally [21]. The role of specific adherence in stable colonization of the human bladder by E. coli has been examined in several studies. In support of the attachment-competition model, E. coli was reportedly capable of adherence, albeit at a moderate level, to exfoliated uroepithelial cells and urinary catheters [22]. However, others suggested that specific bacterial adherence was not required for bladder colonization [8]. A subsequent report indicated that P fimbriae were not required for bladder colonization since a P negative null mutant of E. coli exhibited the same colonization attributes as wild type [23]. Escherichia coli derivatives that contain P-fimbrial genes on a plasmid exhibited enhanced early colonization of the bladder but the P adherence phenotype was lost shortly thereafter [14]. Genes for other E. coli fimbriae classes associated with specific adherence in the urinary tract, including type 1, type 1c, S, Dr, and M adhesions, are either not present in E. coli or are defective [19, 24]. Recent DNA sequence analysis of E. coli revealed a fimbrial gene cluster similar to an uroepithelial cell adhesin called Uca that is found in Proteus mirabilis [25]. Biofilm Prevention Escherichia coli shows a low but detectable level of adherence to silicone urinary catheters. Pretreatment of urinary catheters with E. coli reportedly prevented colonization of the catheters in vitro after challenge with a variety of uropathogens [26]. These results suggest that presence of an E. coli biofilm interferes with subsequent colonization of catheters by pathogenic bacteria. Enhancement of the capacity of E. coli to form a biofilm on catheter material by adding type 1 fimbriae was associated with improvement in bacterial interference in a dose-responsive manner [27]. Although these findings are consistent with the hypothesis that E. coli biofilm presents a physical barrier to attachment by pathogenic bacteria, other interpretations remain possible. Lactobacillus species The topic of potential mechanisms for the therapeutic effect of probiotic Lactobacillus at sites other than the urinary tract has been the subject of several recent reviews [28 30]. Many of the mechanisms that have been suggested to spear head the activity of Lactobacillus probiotics against organisms and infections outside the urinary tract may also apply to prevention of UTI. Potential mechanisms by which Lactobacillus species cause bacterial interference are described below. Immunomodulation Recent studies suggest that modulation of host immune functions may contribute to the therapeutic effect of probiotic Lactobacillus [31]. The authors of a 2-patient pilot study suggested that immunomodulation by Lactobacillus-induced downregulation of proinflammatory cytokines such as tumor necrosis factor α, interleukin 6, interleukin 8, interleukin 10, and interleukin 12 ( p70) may be beneficial [32]. However, factors that were not assessed in this study, such as the presence/absence of bacteriuria during the 90-day trial period, could have confounded the interpretation of results. Gene Regulation Increasing evidence indicates that bacterial products from probiotic Lactobacillus may induce downregulation in expression of virulence factors [30, 33]. Spent culture supernatants from probiotic Lactobacillus reportedly downregulated the promoter activity of E. coli type 1 and P fimbriae genes [34]. Type 1 and P fimbriae are associated with urinary tract colonization and infection, but their role in vaginal colonization has not been investigated. The extent to which Lactobacillus products can prevent specific adherence of uropathogens to the vaginal epithelium or suppress subsequent expression of fimbrial genes after the bacteria migrate to the urinary tract is unknown. Type 1 fimbriae also contribute to colonization of urinary catheters [27]. Additional studies are required to show whether probiotic Lactobacillus can suppress bacterial colonization of urinary catheters in vivo by use of a gene regulation mechanism. Biofilm Prevention Biofilms formed by pathogenic bacteria can facilitate bacterial colonization of tissues and complicate management of 1402 CID 2012:55 (15 November) HEALTHCARE EPIDEMIOLOGY

4 established infection. It is theoretically possible that introduction of nonpathogenic organisms could displace biofilmembedded pathogens in the vagina and, subsequently, prevent UTI. A recent in vitro study showed that application of probiotic L. rhamnosis GR-1 to biofilms formed in vitro by a uropathogenic E. coli strain interfered with the integrity of the biofilm and caused significant killing of E. coli [35]. Production of Antibacterial Agents The mechanism by which Lactobacillus species kill bacterial pathogens could be multifactorial and may include production of bacteriocin-like molecules, hydrogen peroxide, and lactic acid. In a dose- and ph-dependent manner, lactic acid, hydrogen peroxide, and spent culture supernatants from urogenital probiotic strains L. casei var rhamnosus GR-1 and L. reuteri RC-14 all strongly inhibited growth of uropathogenic E. coli strains and increased the promotor activity of outer membrane proteins A and X (porins that are normally upregulated in response to stress by uropathogens) [34]. That same study also showed that lactic acid and Lactobacillus spent culture supernatants could downregulate the promoter activity of the major subunits of type 1 and P fimbriae virulence factors. In another study, hydrogen peroxide acted individually and in cooperation with lactic acid to kill uropathogenic E. coli [36]. Other Mechanisms Probiotic mechanisms for vaginal Lactobacilli that prevent colonization by vaginal pathogens may also prevent vaginal colonization with uropathogens. These include bacteriocin production and aggregation. Whereas both mechanisms have been associated with probiotic Lactobacillus at other bodily sites, their role in UTI prophylaxis has not been investigated. CLINICAL TRIALS The attributes and results of important clinical trials of bacterial interference using E. coli [9, 10, 15, 16, 37 39] or Lactobacillus species [40 45] are summarized in Tables 1 and 2, respectively. Neither E. coli nor Lactobacillus species have reportedly caused symptomatic UTI. Not unexpectedly, all listed studies were relatively small in size ( 100 patients), particularly the E. coli based trials as they required either direct bladder inoculation of nonpathogenic bacteria twice a day for 3 consecutive days (1 set of inoculation) up to 3 sets, or immersion of a bladder catheter in a bacterial suspension of nonpathogenic E. coli for 24 hours before inserting the E. coli coated catheter into the bladder. As a corollary, the finding of an insignificant trend of efficacy in some studies could be attributed to their insufficient power. Although all listed trials were pilot studies intended to generate preliminary data before embarking on pivotal comparative trials, some trials significantly reduced the incidence of UTI by several-fold. Whereas many listed studies were randomized, placebo-led, doubleblind studies, others suffered from a suboptimal single-arm design that dictated comparison of generated data to historic. Recent Infectious Diseases Society of America guidelines on UTI [46] utilized a systematic weighting of the quality of evidence that can be strong (I), moderate (II), or weak (III). Although this review is not intended to provide clinical recommendations, it would be reasonable to use that system to partially assess the quality of evidence emanating from particular subcategories of studies that share the same interfering agent and mode of intervention. Although 2 of the 5 studies that assessed bladder inoculation by E. coli were randomized led and demonstrated significant efficacy [9, 15], the small number of evaluable patients disallows a high quality of evidence. The 2 studies [38, 39] that assessed insertion of an E. coli treated catheter were open-label and very small, and, therefore, generated evidence that would be of moderate quality at best. Although all 4 randomized led studies [40, 43 45] that evaluated Lactobacillus-containing vaginal suppository indicated some level of efficacy, they were not optimally conducted, thereby minimizing the quality of evidence to moderate. The single randomized led trial [42] of an oral drink that contains Lactobacillus is the largest published study on bacterial interference, but other studies need to be done before establishing a high quality of evidence. COMPARISON OF BACTERIAL APPLICATIONS Table 3 compares the attributes of E. coli based vs Lactobacillus-based applications. In contrast to E. coli that could be inoculated into the bladder either directly or via a catheter covered by a biofilm containing E. coli, Lactobacillus species could be introduced in the form of a vaginal suppository or an oral drink. Because the protective effect by nonpathogenic E. coli is based on ling pathogens at the site of infection, the achieved urinary concentrations of nonpathogenic strains of E. coli are usually the same or higher than those of other bacteria that may coexist in the bladder [15]. A different mode of action, of the source of pathogens, applies to Lactobacillus species which replace virulent vaginal flora that could migrate from the vagina to the bladder and cause symptomatic UTI in women. Although intentional bladder colonization with nonpathogenic E. coli strains has been reported to cause local inflammatory symptoms (bladder irritability) and signs (pyuria), we know of no similar issues in patients who receive Lactobacillus-containing supplements. There is no evidence that bacterial interference adversely affects kidney function. Notwithstanding the impact of differences in study design and size on outcome, the protective efficacy of intentional bladder HEALTHCARE EPIDEMIOLOGY CID 2012:55 (15 November) 1403

5 1404 CID 2012:55 (15 November) HEALTHCARE EPIDEMIOLOGY Table 1. Ref Clinical Trials Using Escherichia coli for Prevention of Urinary Tract Infection Author/Year 9 Darouiche 2011 E. coli Strain HU2117 Study Design Randomized, placeboled, 37 Sundén Crossover, placeboled, 38 Prasad Open-label, compared E. coli colonizers to historic 39 Trautner 2007 HU2117 Open-label, compared E. coli recipients to historic 15 Darouiche Darouiche Randomized, placeboled, Open-label, compared E. coli colonizers to noncolonizers and historic 10 Hull Open-label, compared E. coli colonizers to noncolonizers and historic Enrolled Patients Intervention Reported Safety a Reported Efficacy b 65 Bladder inoculation Yes Average no. of episodes of symptomatic UTI per patient-year was lower in the experimental group than in the group (0.50 vs 1.68, P =.02). 20 Bladder inoculation Yes The time to first symptomatic UTI was longer with than without E. coli bacteriuria (median 11.3 vs 5.7 mo, sign test P =.013). There were fewer reports during 1 year of symptomatic UTI with than without E. coli bacteriuria (13 vs 35 episodes; paired t test, P =.009, 95% CI: ). 13 Insertion of a coated bladder catheter for 3 d in patients practicing intermittent bladder catheterization 12 Insertion of a coated bladder catheter for 28 d in patients with indwelling bladder catheters Yes Yes Lower rate of symptomatic UTI while colonized with E. coli than during prestudy period (0.77 vs 2.27 episodes per patient-year). Statistical comparison not provided. Calculated rate of symptomatic UTI in the experimental group was lower than the reported historic rate in spinal cord injured subjects with indwelling bladder catheters (0.15 vs 2.72 cases per 100 patient-days). Statistical comparison not provided. 27 Bladder inoculation Yes Mean number of symptomatic episodes of UTI per year was lower in experimental group than in group (1.6 vs 3.5, P =.036). 44 Bladder inoculation Yes A lower mean rate of symptomatic UTI in patients colonized with E. coli vs patients who could not be colonized (0.06 vs 1.80 episodes of UTI/patient-year, P <.001). 21 Bladder inoculation Yes Mean rates of symptomatic UTI per patientyear were 0 in patients colonized with E. coli vs 3.1 in the same patients before colonization. Statistical comparison not provided. ownloaded from by guest on 06 November 2018 Abbreviations: CI, confidence interval; E. coli, Escherichia coli; UTI, urinary tract infection. a Safety: Intervention is considered safe if inoculated E. coli strain does not cause symptomatic UTI. b Efficacy: Intervention is considered efficacious if bladder colonization by the inoculated E. coli strain significantly reduces the incidence of symptomatic UTI as compared with persons or periods without bacteriuria due to the inoculated strain of E. coli.

6 HEALTHCARE EPIDEMIOLOGY CID 2012:55 (15 November) 1405 Table 2. Ref Clinical Trials Using Lactobacillus species for Prevention of Urinary Tract Infection Author/Year Lactobacillus Strain Study Design 40 Stapleton 2011 L. crispatus CTV05 Randomized, placeboled, 41 Uehara 2006 L. crispatus c Open-label, compared to historic 42 Kontiokari 2001 L. casei var rhamnosus GG 43 Reid 1995 L. casei var rhamnosus GRI & L. fermentum B54 44 Baerheim 1994 L. casei var rhamnosus c 45 Reid 1992 L. casei var rhamnosus GRI & L. fermentum B54 Randomized, led, compared to cranberrylingonberry drink or no drink Randomized, led, compared to Lactobacillus growth factor and historic Randomized, placeboled, Randomized, placeboled, blinded Abbreviations: CI, confidence interval; RR, relative risk; UTI, urinary tract infection. Enrolled Patients a Safety: Intervention is considered safe if the introduced Lactobacillus strain does not cause symptomatic UTI. Intervention Reported Safety a Reported Efficacy b 100 Vaginal suppository Yes Recurrent symptomatic UTI during a 10- week treatment period occurred in 15% of patients in the experimental group vs 27% in the group (RR = 0.5; 95% CI:.2 1.2). 9 Vaginal suppository Yes Average number of recurrent episodes of symptomatic UTI per year was 1.3 in the experimental group vs 5.0 in historic group (P =.0007). 150 Oral drink Yes At 6 mo, the likelihood of recurrent symptomatic UTI was lower (P =.023) in cranberry-lingonberry group (16%) than in Lactobacillus group (38%) or group (36%). 55 Vaginal suppository Yes Rates of symptomatic UTI were comparable among recipients of Lactobacillus and recipients of Lactobacillus growth factor (1.6 vs 1.3 episodes per patient-year); in comparison, the historic group had a 6.0 rate of symptomatic UTIs per patient-year (P <.001) 47 Vaginal suppository Yes A 1.41 ratio of the rate of symptomatic UTI in the placebo group vs the experimental group (95% CI: ). 41 Vaginal suppository Yes Recurrent symptomatic UTI occurred in 21% of patients treated with Lactobacillus suppository vs 47% in those who received skim-milk suppository (P =.27). b Efficacy: Intervention is considered efficacious if recipients of Lactobacillus species have significantly lower incidence of symptomatic UTI as compared to. c Uncharacterized Lactobacillus strain. ownloaded from by guest on 06 November 2018

7 Table 3. Comparison of Approaches Using Escherichia coli vs Lactobacillus Species for Prevention of Urinary Tract Infection Attribute E. coli Lactobacillus species Site/method of inoculation Direct bladder inoculation or coated catheter Vaginal suppository or oral drink Mode of action Controls pathogens at the site of infection Controls the source of pathogens Safety Could cause local inflammatory findings No known issues Efficacy Consistent reduction of urinary tract infection Inconsistent results Practicality Could be cumbersome Easy application Cost Could be high Moderate Coreceipt of antibiotics Antibiotics consistently given before inoculation Inconsistent Applicability Patients with neurogenic or dysfunctional bladder Mostly in healthy women Regulation Food and Drug Administration(FDA) FDA vs dietary supplement colonization with nonpathogenic E. coli strains seems to be more consistent than that afforded by introducing probiotic Lactobacillus. On the other hand, the practicality and cost of intervention favor the Lactobacillus-based approach, which is less invasive and expensive than bladder inoculation with nonpathogenic E. coli strains. As compared to direct bladder inoculation, the insertion of E. coli coated bladder catheter is more practical and less costly. Routine administration of systemic antibiotics before bladder inoculation with E. coli or HU2117 with the aim of reducing the urinary concentration of preexisting bacteria not only augments the likelihood that the bladder will become colonized by these nonpathogenic strains, but also could contribute to higher urinary concentrations of inoculated E. coli strains vs co-colonizing bacteria [9, 15]. Systemic antibiotics, however, are inconsistently given in patients receiving probiotic Lactobacillus. Although bacterial interference is generally intended to improve the quality of life of patients who suffer from recurrent episodes of symptomatic UTI, these 2 particular approaches of bacterial interference apply to distinctly different patient populations; whereas nonpathogenic strains of E. coli have been utilized to prevent symptomatic UTI in patients with neurogenic/dysfunctional bladder (including those with spinal cord injury), the Lactobacillusbased strategy is applied mostly in healthy women. Finally, the use of nonpathogenic strains of E. coli for prevention of UTI requires approval from the Food and Drug Administration (FDA) through an Investigational New Drug application. Although current use of probiotic Lactobacillus for prevention or treatment of infection also requires FDA approval, it is, unfortunately, sometimes used as a dietary supplement to escape the stringent FDA evaluation. FUTURE DIRECTIONS No person is immune from developing UTI. In general, about a third of women have their first UTI by the age of 24 years, and more than a quarter of adult females with a first episode of UTI will have a recurrence [4]. The statistics are even worse in patients who suffer from neurogenic bladder, especially those with spinal cord injury in whom UTI is the most common infectious condition [16]. Preliminary data from the summarized pilot studies are promising, but in and by themselves are not sufficient to establish new standard-of-care practices. Although clinically protective against UTI, limitations to wide-scale implementation of bacterial interference using nonpathogenic strains of E. coli in patients with neurogenic bladder include relative impracticality and high cost of direct bladder inoculation, the possibility that nonpathogenic strains of E. coli could be inflammatory in patients with sensate bladder, and the small size of completed pilot studies. That is why we plan to conduct a randomized, placebo-led, ed, multicenter trial that assesses the efficacy, safety, and economic implications of inserting a bladder catheter that contains a biofilm of nonpathogenic E. coli. As to Lactobacillus species, further work needs to be done to document the stability of probiotic products, assess the effects of specific strains, and explore the impact of probiotic concentration on efficacy. Note Potential conflicts of interest. Both authors are associated with Baylor College of Medicine, which has received grants from ConjuGon. R. O. D. is associated with the Multidisciplinary Alliance Against Device-Related Infections, which has also received grants from ConjuGon. Both authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed. References 1. Vitetta L, Briskey D, Hayes E, Ching C, Peake J. A review of the pharmacobiotic regulation of gastrointestinal inflammation by probiotics, commensal bacteria and prebiotics [published online ahead of print 18 Mar 2012]. Inflammopharmacol Available at: ncbi.nlm.nih.gov/pubmed/ Accessed 1 August Deshpande G, Rao S, Patole S. Progress in the field of probiotics: year Curr Opin Gastroenterol 2011; 27: Hatakka K, Saxelin M. Probiotics in intestinal and non-intestinal diseases. Curr Pharm Design 2008; 14: Barrons R, Tassone D. Use of Lactobacillus probiotics for bacterial genitourinary infections in women: a review. Clin Therapeutics 2008; 30: CID 2012:55 (15 November) HEALTHCARE EPIDEMIOLOGY

8 5. Hooton TM, Bradley SF, Cardenas DD, et al. Diagnosis, prevention, and treatment of catheter-associated urinary tract infection in adults: 2009 International Clinical Practice Guidelines from the Infectious Diseases Society of America. Clin Infect Dis 2010; 50: Hansson S, Jodal U, Lincoln K, Svanborg-Eden C. Untreated asymptomatic bacteriuria in girls: II Effect of phenoxymethylpenicillin and erythromycin given for intercurrent infections. BMJ 1989; 298: Hansson S, Jodal U, Noren L, Bjure J. Untreated bacteriuria in asymptomatic girls with renal scarring. Pediatrics 1989; 84: Andersson P, Engberg I, Lidin-Janson G, et al. Persistence of Escherichia coli bacteriuria is not determined by bacterial adherence. Infect Immun 1991; 59: Darouiche RO, Green BG, Donovan WH, et al. Multicenter randomized led trial of bacterial interference for prevention of urinary tract infection in patients with neurogenic bladder. Urol 2011; 78: Hull R, Rudy D, Donovan W, et al. Urinary tract infection prophylaxis using Escherichia coli in spinal cord injured patients. J Urol 2000; 163: Darouiche RO, Riosa S, Hull RA. Comparison of Escherichia coli strains as agents for bacterial interference. Infect Control Hosp Epidemiol 2010; 31: Kim TK, Thomas SM, Ho M, et al. Heterogeneity of vaginal microbial communities within individuals. J Clin Microbiol 2009; 47: Reid G, Bruce AW. Probiotics to prevent urinary tract infections: the rationale and evidence. World J Urol 2006; 24: Wullt B, Bergsten G, Connell H, et al. P fimbriae enhance the early establishment of Escherichia coli in the human urinary tract. Mol Microbiol 2000; 38: Darouiche RO, Thornby JI, Cerra-Stewart C, Donovan WH, Hull RA. Bacterial interference for prevention of urinary tract infection: a prospective, randomized, placebo-led, pilot trial. Clin Infect Dis 2005; 41: Darouiche RO, Donovan WH, Del Terzo M, Thornby JI, Rudy DC, Hull RA. Pilot trial of bacterial interference for preventing urinary tract infection. Urology 2001; 58: Wullt B, Connell H, Röllano P, Månsson W, Colleen S, Svanborg C. Urodynamic factors influence the duration of Escherichia coli bacteriuria in deliberately colonized cases. J Urol 1998; 159: Trautner BW, Hull RA, Darouiche RO. Colicins prevent colonization of urinary catheters. J Antimicrob Chemother 2005; 56: Zdziarski J, Brzuszkiewicz E, Wullt B, et al. Host imprints on bacterial genomes rapid, divergent evolution in individual patients. PLoS Pathog 2010; 6:e Schierack P, Kleta S, Tedin K, et al. E. coli Nissle 1917 affects Salmonella adhesion to porcine intestinal epithelial cells. PLoS One 2011; 6:e Storm DW, Patel AS, Koff SA, Justice SS. Novel management of urinary tract infections. Curr Opin Urol 2011; 21: Hull RA, Rudy DC, Donovan WH, Wieser IE, Stewart C, Darouiche RO. Virulence properties of Escherichia coli 83972, a prototype strain associated with asymptomatic bacteriuria. Infect Immun 1999; 67: Hull RA, Donovan WH, Del Terzo M, Stewart C, Rogers M, Darouiche RO. Role of type 1 fimbria- and P fimbria-specific adherence in colonization of the neurogenic human bladder by Escherichia coli. Infect Immun 2002; 70: Roos V, Schembri MA, Ulett GC, Klemm P. Asymptomatic bacteriuria Escherichia coli strain carries mutations in the foc locus and is unable to express F1C fimbriae. Microbiology 2006; 152: Cook SW, Mody N, Valle J, Hull R. Molecular cloning of Proteus mirabilis uroepithelial cell adherence (uca) genes. Infect Immun 1995; 63: Trautner BW, Hull RA, Darouiche RO. Escherichia coli inhibits catheter adherence by a broad spectrum of uropathogens. Urol 2003; 61: Trautner BW, Cevallos ME, Li H, et al. Increased expression of type-1 fimbriae by nonpathogenic Escherichia coli results in an increased capacity for catheter adherence and bacterial interference. J Infect Dis 2008; 198: Preidis GA, Hill C, Guerrant RL, Ramakrishna BS, Tannock GW, Versalovic J. Probiotics, enteric and diarrheal diseases, and global health. Gastroenterology 2011; 140: Yoon SS, Sun J. Probiotics, nuclear receptor signaling, and antiinflammatory pathways. Gastroenterol Res Pract 2011; 2011: Reid G, Younes JA, Van der Mei HC, Gloor GB, Knight R, Busscher HJ. Microbiota restoration: natural and supplemented recovery of human microbial communities. Nat Rev Microbiol 2011; 9: Day AS, Keenan JI. Probiotic-mediated modulation of host inflammation. Expert Rev Gastroenterol Hepatol 2011; 5: Anukam KC, Hayes K, Summers K, Reid G. Probiotic Lactobacillus rhamnosus GR-1 and Lactobacillus reuteri RC-14 may help downregulate TNF-alpha, IL-6, IL-8, IL-10 and IL-12 (p70) in the neurogenic bladder of spinal cord injured patient with urinary tract infections: a two-case study. Adv Urol 2009; : Li J, Wang W, Xu SX, Magarvey NA, McCormick JK. Lactobacillus reuteri-produced cyclic dipeptides quench agr-mediated expression of toxic shock syndrome toxin-1 in staphylococci. Proc Natl Acad Sci USA2011; 108: Cadieux PA, Burton JP, Devillard E, Reid G. Lactobacillus by-products inhibit the growth and virulence of uropathogenic Escherichia coli. J Physiol Pharmacol 2009; 60(suppl 6): McMillan A, Dell M, Zellar MP, et al. Disruption of urogenital biofilms by lactobacilli. Colloids Surf B: Biointerfaces 2011; 86: Atassi F, Servin AL. Individual and co-operative roles of lactic acid and hydrogen peroxide in the killing activity of enteric strain Lactobacillus johnsonii NCC933 and vaginal strain Lactobacillus gasseri KS120.1 against enteric, uropathogenic and vaginosis-associated pathogens. FEMS Microbiol Lett 2010; 304: Sundén F, Håkansson L, Ljunggren E, Wullt B. Escherichia coli bacteriuria protects against recurrent lower urinary tract infections in patients with incomplete bladder emptying. J Urol 2010; 184: Prasad A, Cevallos ME, Riosa S, Darouiche RO, Trautner BW. A bacterial interference strategy for prevention of UTI in persons practicing intermittent catheterization. Spinal Cord 2009; 47: Trautner BW, Hull RA, Thornby JI, Darouiche RO. Coating urinary catheters with an avirulent strain of Escherichia coli as a means to establish asymptomatic colonization. Infect Control Hosp Epidemiol 2007; 28: Stapleton AE, Au-Yeung M, Hooton TM, et al. Randomized, placeboled phase 2 trial of a Lactobacillus crispatus probiotic given intravaginally for prevention of recurrent urinary tract infection. Clin Infect Dis 2011; 52: Uehara S, Monden K, Nomoto K, Seno Y, Kariyama R, Kumon H. A pilot study evaluating the safety and effectiveness of Lactobacillus vaginal suppositories in patients with recurrent urinary tract infection. Int J Antimicrob Agents 2006; 28(suppl):S Kontiokari T, Sundqvist K, Nuutinen M, Pokka T, Koskela M, Uhari M. Randomised trial of cranberry-lingonberry juice and Lactobacillus GG drink for the prevention of urinary tract infections in women. BMJ 2001; 322: Reid G, Bruce AW, Taylor M. Installation of Lactobacillus and stimulation of indigenous organisms to prevent recurrence of urinary tract infections. Microecol Ther 1995; 23: Baerheim A, Larsen E, Digranes A. Vaginal application of lactobacilli in the prophylaxis of recurrent lower urinary tract infection in women. Scand J Prim Health Care 1994; 12: Reid G, Bruce AW, Taylor M. Influence of three-day antimicrobial therapy and Lactobacillus vaginal suppositories on recurrence of urinary tract infections. Clin Ther 1992; 14: Hooten TM, Bradley SF, Cardenas DD, et al. Diagnosis, prevention, and treatment of catheter-associated urinary tract infection in adults: 2009 International Clinical Practice Guidelines from the Infectious Diseases Society of America. Clin Infect Dis 2010; 50: HEALTHCARE EPIDEMIOLOGY CID 2012:55 (15 November) 1407