Recurrent urinary tract infection in women

International Journal of Antimicrobial Agents 17 (2001) 259 268 www.ischemo.org Recurrent urinary tract infection in women Thomas M. Hooton * Uni ersity of Washington School of Medicine, Harbor iew Medical Center, Madison Clinic, 325 9th A enue, Seattle, WA 98104-2499, USA Abstract Recurrent urinary tract infections (UTI) are common among young healthy women even though they generally have anatomically and physiologically normal urinary tracts. Women with recurrent UTI have an increased susceptibility to vaginal colonization with uropathogens, which is due to a greater propensity for uropathogenic coliforms to adhere to uroepithelial cells. Risk factors for recurrent UTI include sexual intercourse, use of spermicidal products, having a first UTI at an early age, and having a maternal history of UTIs. Inherited factors may be important in some women with recurrent UTI. Many factors thought to predispose to recurrent UTI in women, such as pre- and post-coital voiding patterns, frequency of urination, wiping patterns, and douching have not been proven to be risk factors for UTI. In contrast to the predominantly behavioral risk factors for young women, mechanical and/or physiological factors that affect bladder emptying are most strongly associated with recurrent UTI in healthy postmenopausal women. The management of recurrent UTI is the same as that for sporadic UTI except that the likelihood of infection with an antibiotic resistant uropathogen is higher in women who have received recent antimicrobials. Strategies to prevent recurrent UTI in young women should include education about the association of recurrent UTI with frequency of sexual intercourse and the usage of spermicide-containing products. Continuous or post-coital prophylaxis with low-dose antimicrobials or intermittent self-treatment with antimicrobials have all been demonstrated to be effective in managing recurrent uncomplicated UTIs in women. Estrogen use is very effective in preventing recurrent UTI in post-menopausal women. Exciting new approaches to prevent recurrent UTI include the use of probiotics and vaccines. Further understanding of the pathogenesis of UTI will lead to more effective and safer methods to prevent these frequent infections. 2001 Elsevier Science B.V and International Society of Chemotherapy. All rights reserved. Keywords: Recurrent UTI; Vaginal colonization; Preventive measures 1. Background A recurrent urinary tract infection (UTI) is a symptomatic UTI that follows clinical resolution of an earlier UTI generally, but not necessarily, after treatment. Recurrent UTIs are common among young healthy women even though they generally have anatomically and physiologically normal urinary tracts. Mabeck found that nearly one-half of the women whose uncomplicated UTIs resolved spontaneously developed a recurrent UTI within the first year [1]. In a recent study of college women with their first UTI, 27% experienced at least one culture-confirmed recurrence within the 6 months following the initial infection [2] and 2.7% had a second recurrence over this time period. In a Finnish study of women aged 17 82 years who had Escherichia * Tel.: +1-206-7315147; fax: +1-206-7315109. E-mail address: hooton@u.washington.edu (T.M. Hooton). coli cystitis, 44% had a recurrence within 1 year, 53% in women older than 55 years and 36% in younger women [3]. Although epidemiologic data for older women are sparse, it is estimated that 10 15% of women over age 60 have frequent recurrences [4]. No large populationbased studies have yet been performed to determine what proportion of women with UTI develop a pattern of high frequency recurrences. Occasionally, recurrences are due to a persistent focus of infection (relapse), but the vast majority is thought to represent re-infection. Thus, among 464 episodes of acute uncomplicated cystitis in college women treated with a variety of antimicrobials, we have found that only 26 (5.6%) had persistent infection, defined as persistence of the initially infecting species within 1 week of the start of therapy [5 7]. In another study of 49 patients, mostly women, some of whom had complicating factors, 84% of recurrences were episodes of re-infection (as determined by biotyping), 0924-8579/01/$ - $20 2001 Elsevier Science B.V and International Society of Chemotherapy. All rights reserved. PII: S0924-8579(00)00350-2

260 T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 typically months apart [8]. Prospective studies have demonstrated that E. coli strains causing UTI may, although appropriately treated and not found in repeated urine cultures in between, cause a new UTI up to 3 years later [9]. It is not clear whether such strains persist in the fecal flora or are reacquired from another source. In a recent study of 23 women with recurrent UTI and 35 women with first episode UTI, E. coli strains were evaluated by chromosomal restriction fragment length polymorphism (RFLP) analysis using pulsed field gel electrophoresis [10]. About 30 (68%) of 44 recurrent UTIs, including several that occurred several months apart, were caused by a strain earlier identified in that individual, while 32 (91%) of the 35 strains from first episode UTIs had unique RFLP profiles. The fecal flora was found to be a reservoir for the strain re-infecting the urinary tract in a subset of patients analyzed. It appears, therefore, that the vast majority of recurrences of cystitis are reinfections and that the initially infecting strain can persist in the fecal flora after elimination from the urinary tract, subsequently re-colonize the introitus and bladder and cause recurrent UTI. It has also been hypothesized that infecting strains may resist clearance in the face of host defenses within the bladder by invading into the epithelium, and later cause recurrent UTI [11]. It is useful to try to distinguish clinically between relapse and recurrence because of urologic evaluation and therapeutic implications. Thus, a recurrence is considered to be a re-infection if it is caused by a different strain from that causing the original infection. It is generally not possible, however, to distinguish between relapse and re-infection with the originally infecting strain. Thus, a recurrence is often defined clinically as a relapse if it is caused by the same species as that causing the original UTI, and if it occurs within 2 weeks after treatment. It is considered a re-infection if it occurs more than 2 weeks after treatment of the original UTI. An exception would be the situation, where a post-treatment urine culture has been investigated and produced no growth of the uropathogen in which case any subsequent recurrence is a re-infection. Most recurrences appear to occur in the first 3 months after the initial UTI [12,13], although clustering has not been found in some studies [3]. 2. Pathogenesis Most uncomplicated UTIs in women cannot be explained by underlying functional or anatomic abnormalities of the urinary tract, but instead appear to result from the interaction of infecting E. coli strains with the woman s epithelial cells. In the healthy person, most uropathogens originate in the person s rectal flora and enter the bladder via the urethra with an interim phase of periurethral and distal urethral colonization. Many host genetic, biologic, and behavioral factors, several of which are discussed below, appear to predispose young healthy women to uncomplicated UTI [14]. Local changes in the vagina, such as ph [15,16] and cervicovaginal antibody [17], and both urine and bladder defense mechanisms [18 22] may play an as yet unidentified role in predisposing women to uncomplicated UTI. The host s inflammatory and immunologic responses help determine the clinical consequences of UTI. Women with recurrent UTI have been shown to have an increased susceptibility to vaginal colonization with uropathogens compared with women without a history of recurrent UTI [14,23 26]. Stamey and Sexton demonstrated that gram negative bacilli were present in 56% of the vaginal cultures from nine women with recurrent UTI but only 24% of those from 20 women without a history of recurrent UTI (P=0.0003) [25]. Those with recurrent UTI had significantly higher prevalence of vaginal colonization with E. coli, Enterococcus faecalis, Proteus mirabilis, and Klebsiella. In women with recurrent UTI compared with controls, colonization with gram-negative bacilli was heavier and lasted longer in these studies. At least in part, this difference between women with and without recurrent UTI appears to result from a greater propensity for uropathogenic coliforms to adhere to the uroepithelial cells of recurrently infected women as compared with cells from women without recurrent infection [27 31]. The underlying reason for this difference has not been established although, in some women, this appears to be genetically determined. In this regard, the nonsecretor phenotype [32 34] and the P 1 phenotype [35,36] are over-represented among girls and women with recurrent UTI and recurrent pyelonephritis, respectively. Further, uroepithelial cells from women who are nonsecretors show enhanced adherence of uropathogenic E. coli compared with cells from secretors [37]. Recent data suggest that the biochemical explanation for the increased adherence of E. coli to nonsecretors uroepithelial cells and for their propensity to develop recurrent UTI may be the presence of unique globoseries glycolipid receptors selectively expressed by epithelial cells of nonsecretors that bind uropathogenic E. coli [38]. This is discussed further below. Certain virulence determinants of uropathogens have also been demonstrated to provide a selective advantage to those strains possessing them with regard to colonization and infection [39,40]. For example, colonization with P-fimbriated strains of E. coli is a strong risk factor for acute uncomplicated pyelonephritis. Whether vaginal colonization and subsequent UTI occur, are the result of a dynamic interaction between these host defense characteristics, host behavioral char-

T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 261 acteristics and uropathogen virulence determinants is not known. The pathogenesis of cystitis is less well understood compared with that of pyelonephritis, and there are no bacterial properties that identify cystitogenic E. coli clones or distinguish them from strains that cause acute pyelonephritis, although hemolysin, type 1 fimbriae and the prsg J96 type of P fimbriae may occur more often in acute cystitis strains than in other E. coli strains [39]. 3. Risk factors 3.1. Healthy premenopausal women Sexual intercourse, diaphragm-spermicide use, and a history of recurrent UTI have been recently shown to be strong and independent risk factors for UTI [41]. Even spermicide-coated condom use results in an increased risk of UTI [42,43]. Recent antimicrobial use, which has been shown to adversely effect vaginal flora in animals and humans [44], also is strongly associated with increased risk of UTI [45]. Risk factors specific for recurrent UTI, however, have received relatively little study. In a recent large case-control study of women with and without a history of recurrent UTI, the strongest risk factor for recurrent UTI in a multivariate analysis was the frequency of sexual intercourse [46]. Other risk factors were spermicide use during the past year, having a new sex partner during the past year, having a first UTI at or before 15 years of age, and having a mother with a history of UTIs. As is true of earlier studies, no associations were found between history of recurrent UTI and pre- and post-coital voiding patterns, frequency of urination, delayed voiding habits, wiping patterns, douching, use of hot tubs, frequent use of pantyhose or tights, or body mass index [46]. It should be pointed out, however, that these behavioral factors have never been evaluated in prospective randomized trials. Having a first UTI at an early age and having a mother with a history of UTIs were associated with 2 4-fold increases in risk and were the most strongly associated variables after sexual intercourse [46]. These findings suggest the possibility that inherited factors may be important in some women with recurrent UTI. As noted earlier, it has been shown that women with a history of recurrent UTIs are several times more likely to be nonsecretors of histo-blood group antigens than are women without such a history. Furthermore, uropathogenic E. coli adhere better to uroepithelial cells of women who are nonsecretors than to cells from secretors [32 34,47]. The biochemical explanation for these observations appears to be that unique globoseries glycolipid receptors that bind uropathogenic E. coli are selectively expressed by epithelial cells of nonsecretors, but not secretors. This is presumably as a result of sialylation of the gal-globoside precursor glycolipid, which in secretors is fucosylated and processed to ABH antigens [48]. More recent studies have shown that these glycosphingolipids are also expressed in human kidney tissue and that they bind P-fimbriated E. coli [49]. The interleukin-8 receptor (IL-8R), CXCR1,is another factor with genetic variability that may influence the development of UTI. lnterleukin-8 is an inflammatory cytokine that promotes neutrophil migration across the infected uroepithelial cells [50,51]. Recently, it has been demonstrated that knockout mice lacking CXCR1 were unable to clear bacteria from the kidney and eventually developed bacteremia [52]. In addition, a preliminary analysis of IL-8R expression on the neutrophils of children with a history of recurrent pyelonephritis has demonstrated a defective version of CXCR1, which the authors suggest may explain their susceptibility to recurrent pyelonephritis [52]. Pelvic anatomy may play a role in some women with recurrent UTI. About 100 women with a history of recurrent UTI (cases) and 113 women with no such history (controls) were recently studied to determine whether there were differences in perineal anatomic measurements, post-void residual urine volume, or in urine voiding characteristics [53]. The distance from the urethra to anus was significantly shorter in cases than in controls (4.8 and 5.0 cm, respectively; P=0.03). Among non-spermicide users, after controlling for sexual intercourse frequency, cases were more likely than controls to have a urethra-to-anus distance less than 4.5 cm (odds ratio(or) =5.7; 95% Cl: 2.0, 16.6) (P= 0.0013). No such difference was found among spermicide users. There were no differences between cases and controls in urethral length, post-void urine residual, or urine voiding characteristics (e.g. peak flow rate, time to peak flow). These data suggest that pelvic anatomic differences may have a role in predisposing some young women to recurrent UTI, especially those who do not have exogenous risk factors for UTI. 3.2. Healthy postmenopausal women The reduced levels of estrogenic hormones present after menopause appear to contribute to the occurrence of recurrent UTI in healthy women. In a double blind, placebo-controlled intervention study discussed below, Raz and Stamm demonstrated that topically applied intravaginal estrogen normalized the vaginal flora and markedly reduced the incidence of recurrent UTI [54]. In a more recent case-control study by Raz and Stamm of 149 healthy postmenopausal women with a history of recurrent UTI and 53 controls without a history of UTI, mechanical and/or physiological factors that affect bladder emptying were found to be strongly associ-

262 T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 ated with recurrent UTIs, in contrast to the predominant behavioral risk factors described for premenopausal women [55]. In this study, three urologic factors were found to be strongly associated with recurrent UTI urinary incontinence (41% of cases vs. 9% of controls; P 0.001); presence of a cystocele (19% vs. 0%; P 0.001); and post-voiding residual urine (28% vs. 2%; P=0.00008). Only six case women, however, had a post-void residual 100 ml. Multivariate analysis showed that urinary incontinence (OR 5.79; P= 0.0009), a history of UTI before menopause (OR 4.85;P=0.003), and nonsecretor status (OR 2.9;P= 0.005) (presence of a cystocele or a post-voiding residual urine were excluded because of low frequency in the controls) were most strongly associated with recurrent UTI. The reason that incontinence predisposes to recurrent UTI is not clear, but is possibly due to the anatomical changes resulting from the procedure (hysterectomy commonly) that led to the incontinence [55]. The authors did not directly evaluate the association of reduced estrogenic hormones and recurrent UTI, but noted that estrogen use was relatively infrequent in both cases and controls [55]. Moreover, they were not able to carefully assess the relationship between sexual intercourse and recurrent UTI because of the reluctance of their patients to discuss sexual practices. Interestingly, only 2% of the women with recurrent UTI thought that coitus was related to their UTIs. 4. Treatment Knowledge of the antimicrobial susceptibility profile of uropathogens causing uncomplicated UTIs in the community, if known, should guide therapeutic decisions for the treatment of acute uncomplicated cystitis. E. coli strains isolated from patients with uncomplicated cystitis are usually susceptible to one or more of the commonly used oral agents, although ampicillin and sulfonamides generally should not be used for empiric therapy because approximately one-third or more of isolates demonstrate in vitro resistance [56]. Of greater concern are the reports showing an increasing proportion of uropathogens causing uncomplicated cystitis resistant to trimethoprim and/or trimethoprim-sulfamethoxazole. It now appears that 15 20% or more of E. coli strains causing acute uncomplicated cystitis are now resistant to these agents in several areas of the US [57] and other countries. As a result, there is increasing concern about empiric use of these agents in regions with a high prevalence of resistance. The prevalence of resistance to nitrofurantoin among E. coli is less than 5%, although non-e. coli uropathogens are often resistant. Resistance to the fluoroquinolones remains below 5% in most studies of uropathogenic strains. However, higher rates of fluoroquinolone resistance among uropathogens are increasingly being reported in some countries around the world. Three-day short course regimens are recommended for the treatment of acute uncomplicated cystitis because of better compliance, lower cost, and lower frequency of adverse reactions than with 7 10 day regimens [58]. Several studies and clinical experience have confirmed the effectiveness of 3-day regimens of trimethoprim, trimethoprim-sulfamethoxazole, or a fluoroquinolone for the treatment of acute uncomplicated cystitis and these agents are generally recommended as the drugs of choice for empiric therapy [58]. There is no apparent benefit in extending therapy with trimethoprim-sulfamethoxazole or a fluoroquinolone past 3 days, and adverse reactions are more common in patients treated with longer regimens. In comparison, 3-day regimens with beta-lactams are less effective than 5 or more days of therapy [58]. Higher cure rates generally have been observed with trimethoprim, trimethoprim-sulfamethoxazole, and fluoroquinolones than with beta-lactams regardless of the site of infection and duration of treatment. Gatifloxacin, recently approved in the US, is approved for the treatment of uncomplicated cystitis, but offers no significant advantage over ciprofloxacin, ofloxacin, or levofloxacin. Nitrofurantoin has been in clinical use for several decades, is well tolerated, and generally demonstrates a consistently low level of resistance among E. coli, grampositive cocci (including E. faecalis), and many Gramnegative species. However, nitrofurantoin is inactive against most Proteus species, including P. mirabilis, and some Enterobacter and Klebsiella strains. There are no published trials in adults that demonstrate nitrofurantoin to be as effective in single-dose or 3-day regimens as trimethoprim-sulfamethoxazole or fluoroquinolones, and one study showed that it was less effective than trimethoprim-sulfamethoxazole given for 3 days [59]. In summary, nitrofurantoin is a safe and generally effective agent for the treatment of uncomplicated cystitis, but it should be administered for a minimum of 7 days. Single-dose regimens are somewhat less effective than 3 7 day regimens, even with fluoroquinolones [56,58]. Fosfomycin tromethamine, approved by the Food and Drug Administration for the single-dose treatment of uncomplicated cystitis, is less effective than 7 10 day regimens of ciprofloxacin or trimethoprim-sulfamethoxazole but comparable to a 7-day regimen of nitrofurantoin (fosfomycin package insert). Fosfomycin, however, has not been compared with single-dose or 3-day regimens of these or other agents. Cefixime and cefpodoxime proxetil are broad-spectrum oral cephalosporins that demonstrate in vitro activity against most uropathogens causing uncomplicated cystitis. There are few data, however, confirming their effectiveness in the treatment of cystitis [56].

T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 263 Trimethoprim-sulfamethoxazole in a 3-day regimen is considered the drug of first choice for acute uncomplicated cystitis in guidelines recently published by the Infectious Disease Society of America [58]. Although as effective, the fluoroquinolones are generally not regarded as first line therapy because of their greater expense and concerns regarding the promotion of quinolone resistance. However, one study demonstrated that the overall cost of a short-course fluoroquinolone regimen may be no more than that of older generic agents [59]. Thus, fluoroquinolones are reasonable firstline agents in women who are known or suspected of having antimicrobial resistant organisms or who are allergic or otherwise do not tolerate more conventional regimens, and in areas where resistance to trimethoprim-sulfamethoxazole is over 15 20% [58]. Nitrofurantoin, oral broad-spectrum cephalosporins, or fosfomycin are also reasonable empiric choices for mild cystitis. Given the increasing prevalence of TMP-SMX resistance among uropathogens, it is important to examine risk factors which predict in vitro resistance. One recent study of UTIs in patients presenting to emergency departments showed that the most important independent risk factors for TMP-SMX resistance were diabetes, recent hospitalization, recent antibiotic use and recent TMP-SMX use [60]. When recently hospitalized patients were excluded from the analyses, only the antibiotic variables remained associated with TMP- SMX resistance. Therefore, it seems reasonable to use TMP-SMX for acute cystitis if the woman has no history of allergy to the drug, has not been on antibiotics in the past 3 6 months (for any reason), and if the TMP-SMX resistance prevalence is not known to be more than 15%. Recurrent cystitis, which occurs during or within the first week following treatment suggests possible relapse and should be managed with a pretreatment urine culture, antimicrobial susceptibility testing and treatment with a fluoroquinolone for 7 days. There is no reason outside of a research study to perform sophisticated typing of initial and recurrent cystitis strains since one cannot eliminate the possibility of recurrence even if the strains are identical. Later recurrences of cystitis can be managed with short course regimens of antimicrobials as described above for sporadic infections. 5. Prevention strategies Sexually active women and those who use spermicides, particularly when used in conjunction with diaphragms, who have frequent recurrent UTI should be counseled about the possible association between their infections and their contraceptive method. Abstinence or a decrease or elimination of the usage of spermicidecontaining products would be expected to reduce the risk of UTI. Cranberry products are widely used for prevention of UTI but their efficacy has yet to be clearly demonstrated. It is reasonable to suggest to women that early post-coital voiding and more liberal fluid intake with the goal of increasing the frequency of micturition, while not shown in case control studies to be associated with a reduced risk of recurrent UTI, might be helpful. Women with frequent recurrent UTIs who do not wish to change their method of contraception should be offered antimicrobial management. 5.1. Antimicrobial management Antimicrobial prophylaxis has been demonstrated to be highly effective in reducing the risk of recurrent UTI in women [61,62]. Prophylaxis has been advocated for women who experience two or more symptomatic UTIs over a 6-month period [62], or three or more over a 12-month period [63], after any existing infection has been eradicated as demonstrated with a negative urine culture 1 2 weeks after treatment. The most important indicator of whether antimicrobial prophylaxis should be tried is the degree of discomfort experienced by the woman as a result of her infections. There are several alternative approaches to choose from and selection of the optimal method of antimicrobial prophylaxis is based on the frequency and pattern of recurrences. Continuous prophylaxis, post-coital prophylaxis, and intermittent self-treatment (which is not really a prophylaxis method) have all been demonstrated to be effective in the management of recurrent uncomplicated UTIs. Regimens for continuous and post-coital prophylaxis are shown in Table 1. Continuous prophylaxis has been demonstrated in numerous studies in different populations to decrease recurrences by 95% when compared with placebo or with patients prior experience (from 2.0 3.0 episodes per patient-year to 0.1 0.2 per patient-year) [62,64,65]. Most authorities advocate a 6-month trial of prophylaxis, with the dose administered at night, after which the regimen is discontinued and the patient observed for further infection. The rationale for the 6-month prophylaxis period is empiric, based on observations that UTIs seem to cluster in some women [12,13]. However, it appears that most women revert back to the earlier pattern of recurrent infections once prophylaxis is stopped unless other factors, such as sexual activity or diaphragm-spermicide use, are modified. Some authorities advocate a longer period of prophylaxis 2 or more years in women who continue to have symptomatic infections [62] and use of trimethoprim-sulfamethoxazole or other agents for as long as 5 years has been reported to be effective and well tolerated [13,66].

264 T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 Table 1 Antimicrobial prophylaxis regimens for women with recurrent urinary tract infection Regimens Continuous Prophylaxis Trimethoprim-sulfamethoxazole 40/200 mg 0 0.2 daily Trimethoprim-sulfamethoxazole 40/200 mg 0.1 thrice weekly Trimethoprim 100 mg daily 0 1.5 b Nitrofurantoin 50 mg daily 0 0.6 Nitrofurantoin 100 mg daily 0 0.7 Cefaclor 250 mg daily 0 Cephalexin 125 mg daily 0.1 Cephalexin 250 mg daily 0.2 Norfloxacin 200 mg daily 0 Ciprofloxacin 125 mg daily 0 Post-coital prophylaxis Trimethoprim-sulfamethoxazole 40/200 mg 0.3 Trimethoprim-sulfamethoxazole 80/400 mg 0 Nitrofurantoin 50 or 100 mg 0.1 Cephalexin 250 mg 0.03 Cinoxacin 250 mg 0.4 Ciprofloxacin 125 mg 0 Norfloxacin 200 mg 0 Ofloxacin 100 mg 0.06 Urinary tract infections per year a a Data from references [62,65,68 71,100,101]. b High recurrence rates observed with trimethoprim associated with trimethoprim-resistance. Post-coital prophylaxis may be a more efficient and acceptable method of prevention than continuous prophylaxis in women whose UTIs appear to be temporally related to sexual intercourse. In the only placebo-controlled trial done to date [67], the infection rate seen with post-coital trimethoprim-sulfamethoxazole (40/200 mg) was 0.3 per patient-year compared with 3.6 per patient-year in placebo-treated women, and the regimen was effective for patients with both low and high intercourse frequencies. These results compare favorably with those observed in other studies with continuous prophylaxis. Uncontrolled studies have demonstrated comparable reduction in infection rates with several other agents used post-coitally [65,68 71]. Depending on the frequency of intercourse, post-coital prophylaxis usually results in considerably less consumption of antimicrobial than does continuous prophylaxis. Some women do not want to take antimicrobials over an extended period of time. If such women are reliable, they may be candidates for self-treatment with a single-dose or 3-day antimicrobial regimen of an antimicrobial such as trimethoprim-sulfamethoxazole or fluoroquinolone. In a randomized crossover trial, 38 women with recurrent UTIs were assigned to use either continuous prophylaxis with trimethoprim-sulfamethoxazole or intermittent self-administered therapy with trimethoprim-sulfamethoxazole (single-dose of 320 mg/1600 mg taken for urinary symptoms) [72]. The infection rate for those on prophylaxis was 0.2 episodes per patient-year compared with 2.2 episodes per patient-year for patients on intermittent self-therapy (P 0.001). In this group of educated women, UTI was correctly diagnosed in 35 (92%) of 38 symptomatic episodes and self-therapy was effective in 30 of the 35 infections. In a more recent study of 34 women, 86% of 78 cultured symptomatic episodes were culture positive, and all the culture positive episodes were cured microbiologically after patient-initiated therapy with a 3-day course of norfloxacin [73]. We have performed similar studies in our student health clinic and have demonstrated similar outcomes using ofloxacin and levofloxacin. Use of this strategy should be restricted to those women who have clearly documented recurrent infections and who are motivated, compliant with medical instructions, and have a good relationship with a medical provider. Such women should be reminded to call their provider if the symptoms are not completely resolved by 48 h. 5.2. Cranberry juice Cranberry juice has been touted as an effective home remedy for preventing UTI for decades, but earlier studies of the effect of cranberry juice on UTI have been small and yielded inconclusive and unimpressive results [74]. One recent case-control study of college women, however, suggested a protective effect against UTI with regular drinking of cranberry juice [75]. Studies have demonstrated that cranberry juice inhibits adherence of E. coli [76] and other gram-negative uropathogens to uroepithelial cells [77]. The urine of humans fed cranberry beverage had significant antiadherence activity as well [76]. The mediators of this anti-adherence effect appear to be fructose that could interfere with adhesion of type 1 fimbriated E. coli to uroepithelium [78] and proanthocyanidins, which can inhibit adherence of P-fimbriated E. coli [79]. Anti-adherence compounds have been isolated from cranberry and blueberry juices, but not from grapefruit, orange, guava, mango, and pineapple juices [80]. Cranberry juice, in contrast to earlier thinking, apparently does not alter the urinary ph [81]. Recently, a placebo-controlled, randomized study was published which supports the efficacy of cranberry juice in preventing asymptomatic bacteriuria in elderly women. In a 6-month study of elderly women (mean age, 78.5 years), 29% of whom were institutionalized, monthly urine samples of 60 women who drank 300 ml of cranberry juice daily were compared with urine samples from 61 women who received a placebo beverage [81]. Bacteriuria with pyuria was found in 28% of

T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 265 samples from the placebo group, but in only 15% of samples from the group who drank cranberry juice (OR=42%; P=0.004). There was also a decrease in the incidence of symptomatic UTI (4 compared with 7%), but this difference did not reach statistical significance. Antibiotic use was decreased by half in the cranberry group. In a more recent small randomized, double-blind, crossover, placebo-controlled clinical trial of 10 women who completed a 6-month study of daily cranberry solids (400 mg) or placebo (3 months on cranberry/3 months placebo), seven had fewer UTIs while on cranberry, two had the same number of UTIs, and one had more UTIs [82]. Of 21 UTIs that occurred while on the study, six occurred during the cranberry months (2.4 per subject year) and 15 occurred during the placebo months (6.0 per subject year) (P 0.005). Although the results of these in vitro and clinical studies are intriguing, and although large numbers of women swear by their cranberry juice, it remains to be determined whether cranberry juice is effective in preventing or treating symptomatic UTI. 5.3. Estrogen use in postmenopausal women Raz and Stamm recently studied 93 postmenopausal women with a history of recurrent UTI in a randomized, double blind, placebo-controlled trial of a topically applied intravaginal estriol cream and evaluated patients monthly for 8 months [54]. Vaginal cultures were obtained at enrollment and after 1 and 8 months. The incidence of UTI in the group given estriol was significantly reduced compared with that in the placebo group (0.5 vs. 5.9 episodes per patient-year, P 0.001). Lactobacilli were absent in all vaginal cultures before treatment and reappeared after 1 month in 61% of the 36 estriol-treated women compared with none of 24 placebo recipients (P 0.001). The prevalence of Enterobacteriaceae fell from 67 to 31% in estriol recipients but was virtually unchanged in the placebo recipients (P 0.005). Likewise, the vaginal ph fell dramatically in the estriol group (mean, 5.5 at entry to 3.6), whereas it remained unchanged in the placebo group. There appeared to be a relation between vaginal colonization with lactobacillus and UTI in that three of 23 estrioltreated women who were colonized with lactobacillus after therapy developed UTI compared with seven of 13 who were not colonized. These data convincingly demonstrate that in postmenopausal women, replacement topical estrogen normalizes the vaginal flora and greatly reduces the risk of UTI. This inverse relationship between vaginal coliform and lactobacillus, particularly hydrogen peroxide producing strains, colonization has been convincingly demonstrated recently in a study of college women [83]. 5.4. Other pre ention strategies Competitive exclusion of uropathogens with lactobacilli or with receptor analogues has been touted as a possible approach to reduce the risk of UTI. Lactobacillus may protect the vagina from colonization by uropathogens through a variety of mechanisms including steric hindrance or blocking potential sites of attachment [84,85], production of hydrogen peroxide, which may prevent vaginal colonization with uropathogens [86,87], production of inhibitors of bacterial growth [88,89], or by maintenance of a low ph [16]. This approach holds promise, but studies to date have been unconvincing. Other potential strategies include the use of a genetically engineered avirulent E. coli strain that could be instilled intravesically in women, which could prove useful in high-risk patients who continually fail to respond to conventional prophylaxis measures [90]. Compounds mimicking the host receptors may competitively bind to bacterial surface ligands, which might decrease the number of bacteria attaching to the mucosa sufficiently to alter the delicate balance of hostparasite interaction in favor of the host [91]. Prevention strategies might also include the development of soluble carbohydrate inhibitors of bacterial adhesion to the urogential mucosa [92]. A safe and effective vaccine to reduce the risk of recurrent UTI would be a welcome breakthrough. Whole cell vaccines, made from combinations of heatkilled uropathogenic strains delivered by injection or by a vaginal suppository, have had to date only partial success with their protective effect diminishing over several weeks [93]. One group pursuing a promising approach is developing a vaccine based on E. coli type 1 fimbrial components [94]. Virtually all uropathogenic strains of E. coli assemble type 1 pili that contain the FimH adhesin. This group has shown that sera from animals vaccinated with candidate FimH vaccines inhibited uropathogenic E. coli from binding to human bladder cells in vitro. Immunization with FimH reduced in vivo colonization of the bladder mucosa by more than 99% in a murine cystitis model. Furthermore, passive systemic administration of immune sera to FimH also resulted in reduced bladder colonization by uropathogenic E. coli. This promising approach is currently undergoing human trials. 6. Urologic evaluation The majority of women with recurrent uncomplicated cystitis has no anatomical or functional abnormality of the urinary tract and, therefore, does not need an evaluation of their urinary tract. Studies of the value of excretory urography and of cystoscopy in women

266 T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 with recurrent UTI have demonstrated that significant abnormalities which influence subsequent management of UTIs are very uncommon [95 98]. Thus, routine evaluation of recurrent cystitis patients with excretory urography or other invasive techniques results in unnecessary expense and potential toxicity. However, further evaluation of the urinary tract is recommended if, with any of the recurrences, suspicion arises as to the presence of complicating factors such as structural or functional abnormalities of the genitourinary tract. Routine urologic investigation of young women with acute pyelonephritis is also generally not cost-effective and has a low diagnostic yield [99]. Although women with recurrent acute pyelonephritis are often evaluated with excretory urography or ultrasound soon after hospitalization, in our experience with otherwise healthy college women it is rare to find urologic abnormalities in those who respond promptly to antimicrobial therapy. We recommend such an evaluation after 2 recurrences of pyelonephritis or if any complicating factor is identified with any of the recurrences. In those women in whom further evaluation is indicated, we recommend starting with computed tomography or a renal ultrasound to rule out nephrolithiasis or obstructive uropathy. A complete urologic evaluation, including cystoscopy and excretory urography, should be performed in patients who have persistent hematuria after the infection has been eradicated. References [1] Mabeck CE. Treatment of uncomplicated urinary tract infection in non-pregnant women. Postgrad Med J 1972;48:69 75. [2] Foxman B. Recurring urinary tract infection: incidence and risk factors. Am J Public Health 1990;80:331 3. [3] Ikaheimo R, Sutonen A, Heiskanen T, Karkkainen U, Kuosmanen P, et al. Recurrence of urinary tract infection in a primary care setting: analysis of a 1-year follow-up of 179 women. Clin Infect Dis 1996;22:91 9. [4] Romano JM, Kaye D. UTI in the elderly: common yet atypical. Geriatrics 1981;36:113 5. [5] Hooton TM, Winter C, Tiu F, Stamm WE. Randomized comparative trial and cost analysis of 3-day antimicrobial regimens for treatment of acute cystitis in women. J Am Med Assoc 1995;273:41 5. [6] Hooton TM, Johnson C, Winter C, et al. Single-dose and three-day regimens of ofloxacin versus trimethoprim-sulfamethoxazole for acute cystitis in women. Antimicrob Agent Chemother 1991;35:1479 83. [7] Hooton TM, Latham RH, Wong ES, Johnson C, Roberts PL, Stamm WE. Ofloxacin versus trimethoprim-sulfamethoxazole for treatment of acute cystitis. Antimicrob Agents Chemother 1989;33:1308 12. [8] McGeachie J. Recurrent infection of the urinary tract: reinfection or recrudescence? Br Med J 1966;1:952 4. [9] Brauner A, Jacobson SH, Kuhn I. Urinary Escherichia coli causing recurrent infections-a prospective follow-up of biochemical phenotypes. Clin Nephrol 1992;38:318 23. [10] Russo TA, Stapleton A, Wenderoth S, Hooton TM, Stamm WE. Chromosomal restriction fragment length polymorphism analysis of Escherichia coli strains causing recurrent urinary tract infections in young women. J Infect Dis 1995;172:440 5. [11] Mulvey MA, Lopez-Boado YS, Wilson CL, et al. Induction and evasion of host defenses by type-1 piliated uropathogenic Escherichia coli. Science 1998;282:1494 7. [12] Kraft JK, Stamey TA. The natural history of symptomatic recurrent bacteriuria in women. Medicine 1977;56:55. [13] Stamm WE, McKevitt M, Roberts PL, White NJ. Natural history of recurrent urinary tract infections in women. Rev Infect Dis 1991;13:77. [14] Stamey TA, Timothy M, Millar M, Mihara G. Recurrent urinary infections in adult women. The role of introital enterobacteria. Calif Med 1971;115:1 19. [15] Stamey TA, Kaufman MF. Studies of introital colonization in women with recurrent urinary infections. II. A comparison of growth in normal vaginal fluid of common versus uncommon serogroups of Escherichia coli. J Urol 1975;114:264 7. [16] Stamey TA, Timothy MM. Studies of introital colonization in women with recurrent urinary infection: I. The role of vaginal ph. J Urol 1975;114:261 3. [17] Stamey TA, Wehner N, Mihara G, Condy M. The immunologic basis of recurrent bacteriuria: role of cervicovaginal antibody in enterobacterial colonization of the introital mucosa. Medicine 1978;57:47 56. [18] Cox CE, Hinman F. Experiments with induced bacteriuria, vesical emptying and bacterial growth on the mechanism of bladder defense to infection. J Urol 1961;86:739 48. [19] Norden CW, Green CM, Kass ER. Antibacterial mechanisms of the urinary bladder. J Clin Invest 1968;47:2689 700. [20] Parsons CL, Greenspan C, Mulholland SG. The primary antibacterial defense mechanism of the bladder. Invest Urol 1975;13:72 6. [21] Mulholland SG. Lower urinary tract antibacterial defense mechanisms. Invest Urol 1979;17:93 7. [22] Orskov I, Ferencz A, Orskov F. Tamm-Horsfall protein or uromucoid is the normal urinary slime that traps type 1 fimbriated Escherichia coli (letter). Lancet 1980;1:887. [23] Pfau A, Sacks T. The bacterial flora of the vaginal vestibule, urethra and vagina in premenopausal women with recurrent urinary tract infections. J Urol 1981;126:630 4. [24] Fowler JE, Latta R, Stamey TA. Studies of introital colonization in women with recurrent urinary infections: VIII. The role of bacterial interference. J Urol 1977;118:296 8. [25] Stamey TA, Sexton CC. The role of vaginal colonization with Enterobacteriaciae in recurrent urinary tract infections. J Urol 1975;113:214 7. [26] Schaeffer AJ, Stamey TA. Studies of introital colonization in women with recurrent urinary infections. IX. The role of antimicrobial therapy. J Urol 1977;118:221 4. [27] Schaeffer AJ, Jones JM, Falkowski WS, Duncan JL, Chmiel JS, Plotkin BJ. Variable adherence of uropathogenic Escherichia coli to epithelial cells from women with recurrent urinary tract infection. J Urol 1982;128:1227 30. [28] Svanborg-Eden C, Jodal U. Attachment of Escherichia coli to urinary sediment epithelial cells from urinary tract infectionprone and healthy children. Infect Immun 1979;26:837 40. [29] Fowler JE, Jr, Stamey TA. Studies of introital colonization in women with recurrent urinary infections. VII. The role of bacterial adherence. J Urol 1977;117:472 6. [30] Kallenius G, Winberg J. Bacterial adherence to periurethral epithelial cells in girls prone to urinary-tract infections. Lancet 1978;2:540 3. [31] Schaeffer AJ, Jones JM, Dunn JK. Association of in vitro Escherichia coli adherence to vaginal and buccal epithelial cells with susceptibility of women to recurrent urinary-tract infections. New Engl J Med 1981;304:1062 6.

T.M. Hooton / International Journal of Antimicrobial Agents 17 (2001) 259 268 267 [32] Hooton TM, Roberts PL, Stamm WE. Effects of recent sexual activity and use of a diaphragm on the vaginal microflora. Clin Infect Dis 1994;19:274 8. [33] Kinane DF, Blackwell CC, Brettle RP, Weir DM, Winstanley FP, Elton RA. ABO blood group, secretor state, and susceptibility to recurrent urinary tract infection in women. Br Med J 1982;285:7 9. [34] Sheinfeld J, Schaeffer AJ, Cordon-Cardo C, Rogatko A, Fair WR. Association of the Lewis blood-group phenotype with recurrent urinary tract infections in women. New Engl J Med 1989;320:773 7. [35] Lomberg H, Hanson LA, Jacobsson B, Jodal U, Leffler H, Svanborg Eden C. Correlation of P blood group, vesicoureteral reflux, and bacterial attachment in patients with recurrent pyelonephritis. New Engl J Med 1983;308:1189 92. [36] Tomisawa S, Kogure T, Kuroume T, et al. P blood group and proneness to urinary tract infection in Japanese children. Scand J Infect Dis 1989;21:403 8. [37] Lomberg H, Cedergren B, Leffler H, Nilsson B, Carlstrom AS, Svanborg-Eden C. Influence of blood group on the availability of receptors for attachment of uropathogenic Escherichia coli. Infect Immun 1986;51:919 26. [38] Stapleton A, Nudelman E, Clausen H, Hakomori S, Stamm WE. Binding of uropathogenic Escherichia coli R45 to glycolipids extracted from vaginal epithelial cells is dependent on histo-blood group secretor status. J Clin Invest 1992;90:965 72. [39] Svanborg C, Godaly G. Bacteral virulence in urinary tract infection. Infect Dis Clin North Am 1997;11:513 29. [40] Johnson JR. Virulence factors in Escherichia coli urinary tract infection. Clin Microbiol Rev 1991;4:80 128. [41] Hooton TM, Scholes D, Hughes JP, Winter C, Roberts PL, et al. A prospective study of risk factors for symptomatic urinary tract infection in young women. New Engl J Med 1996;335:468 74. [42] Fihn SD, Boyko EJ, Normand EH, Chen CL, Grafton JR, et al. Association between use of spermicide-coated condoms and Escherichia coli urinary tract infection in young women. Am J Epidemiol 1996;144:512 20. [43] Fihn SD, Boyko EJ, Chen CL, Normand EH, Yarbro P, et al. Use of spermicide-coated condoms and other risk factors for urinary tract infection caused by Staphylococcus saprophyticus. Arch Intern Med 1998;158:281 7. [44] Hooton TM, Stamm WE. The vaginal flora and UTIs. In: Mobley HLT, Warren JW, editors. UTIs: Molecular Pathogenesis and Clinical Management. Washington, DC: ASM Press, 1996:67 94. [45] Smith HS, Hughes JP, Hooton TM, Roberts P, Scholes D, et al. Antecedent antimicrobial use increases the risk of uncomplicated cystitis in young women. Clin Infect Dis 1997;25:63 8. [46] Scholes D, Hooton TM, Roberts PL, Stapleton AE, Gupta K, Stamm WE. Risk factors for recurrent UTI in young women. J Infect Dis 2000;182:1177 82. [47] Lomberg H, Cedergren B, Leffler H, Nilsson B, Carlstrom AS, et al. Influence of blood group on the availability of receptors for attachment of uropathogenic Escherichia coli. Infect Immun 1986;51:919 26. [48] Stapleton A, Nudelman E, Clausen H, Hakomori S, Stamm WE. Binding of uropathogenic Escherichia coli R45 to glycolipids extracted from vaginal epithelial cells is dependent on histo-blood group secretor status. J Clin Invest 1992;90:965 72. [49] Stapleton AE, Stroud MR, Hakomori SI, Stamm WE. The globoseries glycosphingolipid sialosyl galactosyl globoside is found in urinary tract tissues and is a preferred binding receptor in vitro for uropathogenic Escherichia coli expressing papendoded adhesins. Infect Immun 1998;66:3856 61. [50] Godaly G, Proudfoot AE, Offord RE, Svanborg C, Agace WW. Role of epithelial interleukin-8 (IL-8) and neutrophil IL-8 receptor A in Escherichia coli-induced transuroepithelial neutrophil migration. Infect Immun 1997;65:3451 6. [51] Godaly G, Frendeus B, Proudfoot A, Svensson M, Klemm P, Svanborg C. Role of fimbriae-mediated adherence for neutrophil migration across Escherichia coli-infected epithelial cell layers. Mol Microbiol 1998;30:725 35. [52] Frendeus B, Godaly G, Karpman D, Svanborg C. Receptor deficiency confers susceptibiltiy to acute pyelonephritis. In Program and Abstracts of the Thirty-Ninth lnterscience Conference on Antimicrobial Agents and Chemotherapy, San Francisco, CA,. Abstract LB-S, American Society for Microbiology, Washington, DC, 1999, 52. [53] Hooton TM, Stapleton AE, Roberts PL, Winter C, Scholes D, Bavendam T, et al. Perineal anatomy and urine-voiding characteristics of young women with and without recurrent urinary tract infections. Clin Infect Dis 1999;29:1600 1. [54] Raz R, Stamm WE. A controlled trial of intravaginal estriol in postmenopausal women with recurrent urinary tract infections. New Engl J Med 1993;329:753 6. [55] Raz R, Gennesin Y, Wasser J, Stoler Z, Rosenfeld S, Rottensterich E, Stamm WE. Recurrent urinary tract infection in postmenopausal women. Clin Infect Dis 2000;30:152 6. [56] Hooton TM, Stamm WE. Diagnosis and treatment of uncomplicated urinary tract infection. Infect Dis Clin North Am 1997;11:551 81. [57] Gupta K, Scholes D, Stamm WE. Increasing prevalence of antimicrobial resistance among uropathogens causing acute uncomplicated cystitis. J Am Med Assoc 1999;281:736 8. [58] Warren JW, Abrutyn E, Hebel JR, Johnson JR, Schaeffer AJ, Stamm WE. Guidelines for antimicrobial treatment of uncomplicated acute bacterial cystitis and acute pyelonephritis in women. Clin Infect Dis 1999;29:745 58. [59] Hooton TM, Winter C, Tiu F, Stamm WE. Randomized comparative trial and cost analysis of 3-day antimicrobial regimens for treatment of acute cystitis in women. J Am Med Assoc 1995;273:41. [60] Wright SW, Wrenn KD, Haynes ML. Trimethoprim-sulfamethoxazole resistance among urinary coliform isolates. J Gen Intern Med 1999;14:606 9. [61] Stamm WE, Hooton TM. Management of urinary tract infections in adults. New Engl J Med 1993;329:1328 34. [62] Nicolle LE, Ronald AR. Recurrent urinary tract infection in adult women: diagnosis and treatment. Infect Dis Clin North Am 1987;1:793 806. [63] Ronald AR, Conway B. An approach to urinary tract infections in ambulatory women. Curr Clin Top Infect Dis 1988;9:76 125. [64] Nicolle LE. Prophylaxis: recurrent urinary tract infection in women. Infection 1992;20:5203 5. [65] Chew LD, Fihn SD. Recurrent cystitis in nonpregnant women. West J Med 1999;170:274 7. [66] Nicolle LE, Harding GKM, Thomson M, Kennedy J, Urias B, Ronald AR. Efficacy of five years of continuous, low-dose trimethoprim-sulfamethoxazole prophylaxis for urinary tract infection. J Infect Dis 1988;157:1239 41. [67] Stapleton A, Latham RH, Johnson C, Stamm WE. Postcoital antimicrobial prophylaxis for recurrent urinary tract infection. J Am Med Assoc 1990;264:703 6. [68] Pfau A, Sacks T, Engelstein D. Recurrent urinary tract infection in premenopausal women: prophylaxis based on an understanding of the pathogenesis. J Urol 1983;129:1153. [69] Pfau A, Sacks TG. Effective prophylaxis of recurrent urinary tract infections in premenopausal women by post-coital administration of cephalexin. J Urol 1989;142:1276. [70] Pfau A, Sacks TG. Effective postcoital quinolone prophylaxis of recurrent urinary tract infections in women. J Urol 1994;152:136.