Urinary Tract Infections: Disease Panorama and Challenges [626260]

S1
Urinary Tract Infections: Disease Panorama and Challenges
Walter E. Stamm1and S. Ragnar Norrby2 1Division of Allergy and Infectious Diseases, University of
Washington, Seattle, Washington;2Division of Infectious Diseases,
University of Lund, Lund, Sweden
The major goal of this conference was to review new advances
in our undertaking of the pathogenesis of urinary tract infec-
tions (UTIs) from the perspective of both the pathogen andthe host. This initial presentation provided an appropriate con-text for subsequent discussion by providing an overview of theetiology and epidemiology of UTI, diagnostic and therapeuticapproaches, current management strategies, evolution of anti-microbial resistance, and current approaches to prevention.
Epidemiology
It is estimated that 150 million UTIs occur yearly on a global
basis, resulting in more than 6 billion dollars in direct health
care expenditures [1]. Infections are rare in boys except in as-sociation with anatomic or functional abnormalities in the ®rstyear of life [2]. Infections are also infrequent among 2- to 13-year-old girls, but some young girls experience multiple re-peated episodes of recurrent cystitis or pyelonephritis [2]. Theincidence of UTIs markedly increases among young womenduring adolescence, with an estimated 7 million acute uncom-plicated infections occurring annually in the United States, asdetermined on the basis of data extrapolated from surveys ofphysicians [3]. In a recent prospective study of acute UTIs inyoung women, the incidence was 0.5±0.7 per year [4]. Recurrentinfections become a problem in 25%±30% of women who ex-perience an initial infection. Most uncomplicated infections inwomen do not result in long-term sequelae or renal damage.However, such infections typically cause about 6 days of disa-bility per episode, and thus, in the aggregate, result in sub-stantial morbidity due to their considerable frequency; indeed,they are the most frequent bacterial infections in women [5].Estimates of UTI incidence among young men in the same agerange are several logs less at 5±8 infections per 10,000 [6]. Onthe basis of hospital survey data, at least 250,000 cases of py-elonephritis per year occur in the United States [7].
Among complicated UTIs, those occurring in the hospital as
nosocomial infections (primarily urinary catheter±related in-fections) are the only subgroup for which accurate estimates offrequency are available. Catheter-associated UTIs account for
Grant support: NIH (DK-53369, DK-47549).
Reprints or correspondence: Dr. Walter E. Stamm, University of Wash-
ington, School of Medicine, Box 356523, 1959 NE Paci®c St., Seattle, WA
98195 ([anonimizat]).
The Journal of Infectious Diseases 2001;183(Suppl 1):S1-4
q2001 by the Infectious Diseases Society of America. All rights reserved.
0022-1899/2001/18305S-0001$02.0040% of all hospital-acquired infections and are thus the most
common type of nosocomial infection [8]. Overall, about 10%of patients with short-term catheterization develop infections.This results in an estimated 1±1.5 million catheter-associatedUTIs in the United states per year. Besides accounting for extrahospital costs ( »$400/per episode), these infections may be an
important reservoir for selection and transmission of multi-drug±resistant strains and are a frequent source of gram-neg-ative bacteremia in hospitalized patients [8].
As indicated above, most determinations of UTI incidence
and of the costs and morbidity attributable to UTIs have beenderived from small, often nonrepresentative databases and,thus, could be described as guesstimates. Data are particularlyabsent from developing countries, where it is not knownwhether UTIs are more frequent or severe than in developedcountries. Additional prospective population-based studies ofincidence and cost would be very useful in de®ning where re-search should best be directed. It is not clear, for example,whether UTIs are increasing or decreasing in incidence. Studiesfocused on subgroups of patients who are likely to suffer morefrequent infection or more severe complications of infection(e.g., diabetics or patients with spinal cord injury) would bevery useful.
Etiology
A bacterial etiology can usually be demonstrated for most
episodes of UTI, and the most common bacterial species seen
in various patient groups have now been well de®ned. However,there remain patient groups in whom etiology is unclear. Evenamong young women with acute symptoms and pyuria, thereis subset in whom no pathogens can be isolated. A substantialnumber of men with prostatitis of the nonbacterial but in¯am-matory category have no apparent microorganism demonstra-ble, and interstitial cystitis remains an idiopathic syndrome.Assessment of such patients for newly recognized pathogens,such as Mycoplasma genitalium, or the use of broad host±range
16s rRNA probes to identify new pathogens in these patientgroups would be of interest.
Diagnosis
Although molecular diagnostic approaches utilizing antigen
detection or DNA hybridization and ampli®cation techniques
are being applied to the diagnosis of many infections, UTI arestill generally diagnosed as they have been for decadesÐthat
Downloaded from https://academic.oup.com/jid/article-abstract/183/Supplement_1/S1/2190986
by guest
on 20 March 2018

S2 Stamm and Norrby JID 2001;183 (Suppl 1)
is, by urine culture results. Patient care could bene®t greatly
from the development of a rapid (i.e., patient bedside), accurate,inexpensive test that would allow the practitioner to decidewhom to treat at the time of examination. Such a test wouldhave the potential to reduce unnecessary antimicrobial use andreduce the emergence of resistance. Since UTI can be causedby a variety of bacterial species, an approach targeting speci®cbacterial genes or antigens would likely not be suf®ciently sen-sitive. An alternative approach might be to develop a diagnostictest aimed at recognition of one or more components of theinnate host response, which likely would be positive regardlessof the infecting bacterial species. In one sense, the leukocyteesterase test could be regarded as an early version of this ap-proach, but it has not proven to be suf®ciently sensitive orspeci®c.
Another research need is the development of criteria for as-
sessing the success or failure of treatment for UTI. While muchresearch has been directed toward trying to de®ne optimal col-ony-count criteria for the diagnosis of UTI, almost none hasbeen devoted to criteria for de®ning the cure. While variouscriteria have been proposed, the application of these differentcriteria to the same dataset gives greatly divergent results.
Last, there is an outstanding need for a convenient nonin-
vasive means to distinguish bladder from renal infection. Theantibody-coated bacterial test was the last attempt to accom-plish this, but it lacked both sensitivity and speci®city. Thepresence of speci®c molecules or genes of renal origin in theurine or the presence of speci®c host-response molecules thatare indicators of renal in¯ammation might be possible ap-proaches. Characterization of infectious episodes as renal orbladder infections would lead to much greater precision andnew insights in epidemiologic and treatment studies.
Changing Management Strategies
In an era of health maintenance organizations and cost con-
sciousness, many doctors have advocated that acute uncom-
plicated UTI in women can be managed effectively and safelyby utilizing empiric antibiotic therapy without doing a urineculture or even looking for pyuria. Approaches have rangedfrom diagnosis and care by telephone to requiring an of®ce visitbut no laboratory testing [9]. The rationale for this approachrests primarily upon (1) the narrow and predictable spectrumof agents causing UTIs in young women, (2) the predictableantimicrobial susceptibilities of these agents, (3) the ability ofwomen and health care providers to recognize a UTI on thebasis of clinical symptoms and signs, and (4) the usually ex-cellent response to short-course antibiotic therapy.
Many ăguidelinesș for the treatment of uncomplicated UTIs
in women have been developed and implemented. While thegeneral clinical impression has been that this approach hasreduced the costs of laboratory testing and patient visits andmay have improved antimicrobial use, the overall effectivenessof the approach has been little studied. Potentially negativeoutcomes could include misdiagnosis (and mistreatment) of sex-ually transmitted diseases or pyelonephritis as uncomplicatedUTI, increased UTI recurrences or treatment failures, or misuseof antibiotics for purposes other than UTI treatment. Furthercritical study of the clinical and cost effectiveness of such man-agement guidelines would be very useful for addressing thesepossible negative outcomes.
Another related proposed change in treatment strategy for
women with recurrent UTIs has been the use of patient-initiatedhome therapy [10, 11]. This approach depends on the abilityof the woman who has recurrent UTI to correctly self-diagnosethe UTI and to initiate therapy with an antimicrobial alreadyin her possession. Studies to date suggest this approach worksvery well in selected patients with a history of recurrent UTI.The accuracy of self-diagnosis exceeds 90%, and treatment ef-®cacy is also high.
Treatment
General trends in the treatment of UTIs over the past decade
have been toward the use of shorter regimens (even single-dose
therapy); once-a-day dosing; and in the case of acute pyelo-nephritis and complicated UTI, the provision of therapy in theoutpatient setting (for most patients) with ¯uoroquinolones andother broad-spectrum drugs that have excellent oral absorptionand pharmacokinetics. Evidence-based treatment guidelines foracute uncomplicated UTI (cystitis and pyelonephritis) have re-cently been developed and published by the Infectious DiseasesSociety of America [12].
For cystitis, 3-day regimens of trimethoprim-sulfamethoxa-
zole (TMP-SMX), cipro¯oxacin, or o¯oxacin are recommendedas the most cost- and clinically effective and best-tolerated ap-proach. Single-dose therapy is less effective than 3-day therapy,especially with û-lactam agents. Seven-day regimens with TMP-SMX or ¯uoroquinolones are no more effective than 3-dayregimens, but they result in additional side effects.
For acute pyelonephritis, oral therapy with a ¯uoroquinolone
(generally cipro¯oxacin or o¯oxacin) can be used in many pa-tients who have no nausea or vomiting and no signs of hypo-tension or sepsis. Seven days of therapy is generally suf®cient.Sicker patients may require hospitalization for initial parenteraltherapy for 1±3 days, followed by completion of the regimewith oral therapy.
A similar approach to the treatment of complicated UTI can
be used in the outpatient setting for patients with mild illness.In such cases, well-absorbed, broad-spectrum oral agents (e.g.,¯uoroquinolones) could be used, with initial hospitalization forsicker patients. However, there have been few controlled trialsof complicated UTI addressing either the optimal length oftherapy or the optimal drug (see below).
Downloaded from https://academic.oup.com/jid/article-abstract/183/Supplement_1/S1/2190986
by guest
on 20 March 2018

JID 2001;183 (Suppl 1) UTIs: Disease Panorama and Challenges S3
Antimicrobial Resistance
Studies clearly demonstrate increasing antibiotic resistance
in uropathogens causing both community- and nosocomially
acquired UTIs [13]. Even in women with acute uncomplicatedUTI, increasing resistance to ampicillin (30%±40%), cephalo-thin (20%±30%), and TMP-SMX (15%±20%) has been dem-onstrated in causative Escherichia coli. In the United States and
much of Europe, resistance to nitrofurantoin and ¯uoroquin-olones remains rare in these strains. However, elsewhere in theworld, TMP resistance may be as high as 50%, and ¯uoro-quinolone-resistant E. coli are not uncommon in uncomplicated
UTI. Antimicrobial resistance has also become an increasingproblem in complicated and catheter-associated UTI due toTMP-SMX and ¯uoroquinolone resistance among E. coli and
other enterobacteria as well as to extended-spectrum û-lacta-mase±mediated resistance in gram-negative bacteria and van-comycin-resistance in enterococci.
The impact of in vitro resistance on clinical outcome in UTI
requires further study. Since many drugs achieve very high andprolonged urinary levels, it is possible that in vitro testing doesnot accurately predict clinical outcome. In a study of womenwith acute uncomplicated pyelonephritis, Talan et al. [14]clearly demonstrated that patients who were infected withTMP-SMX±resistant strains and treated with TMP-SMXachieved only a 50% cure, compared with a 90% cure in thosewho received TMP-SMX for TMP-SMX±sensitive strains.With acute uncomplicated cystitis, where high urinary concen-trations in the bladder urine might be expected to overcome invitro resistance, more data are needed to ascertain whetherresistance to TMP-SMX does indeed predict a higher rate oftreatment failure [15].
Given the continuing evolution of drug resistance, there is a
need for ongoing surveillance to accurately monitor trends andrecommend effective empiric treatment regimens. Because em-piric therapy is being widely used, fewer urinary infections arebeing routinely cultured. Thus, those patients who have cultureresults are more likely to re¯ect a selection bias toward com-plications, recent treatment, recurrence, or suspected resistance.Laboratory-based surveillance can overestimate resistanceprevalence for this reason. Patient-based surveillance of anti-microbial resistance using consecutive or randomly selected pa-tients would be a more accurate but more costly approach tosurveillance of patients with speci®c syndromes, such as acutecystitis or acute pyelonephritis. Studies to identify risk factorsor predictors of antimicrobial resistance in patients with UTIare also needed.
Risk Factors for UTI in Women
Considerable research has focused on identifying the pre-
dominant risk factors for acute uncomplicated UTIs in women,
the most common group developing UTI. These studies canbest be summarized by categorizing the identi®ed risk factorsin premenopausal versus postmenopausal women and thoseassociated with a single (or sporadic) infection versus those seenin women with recurrent infection [16]. It is not surprising thatmany of the risk factors seen in sporadic infections are alsoseen in recurrent diseases.
In premenopausal women, coitus, spermicide exposure, prior
history of UTI, and recent antibiotic exposure have been iden-ti®ed as key risk factors for UTI in both case-control and pro-spective studies. Spermicide exposure likely modi®es the normalvaginal ¯ora via a differential antimicrobial effect so as to eradi-cate the normal introital ¯ora and enhance E. coli colonization.
These same factors appear to be important in recurrent infec-tion in young women, as do two additional factors, namely amaternal history of UTI or a history of childhood onset ofUTI. The latter two variables may be consistent with an in-herited predisposition to recurrence, as might nonsecretorstatus (which has been associated with recurrent UTI in severalstudies).
In postmenopausal women, the predominant factors asso-
ciated with recurrent UTI are anatomic or functional defects,such as incontinence, post-void residual urine, or cystocele.Nonsecretor status and prior history of UTI, as in youngerwomen, are both related to recurrent UTI in postmenopausalwomen. Last, the relative lack of estrogen accompanying men-opause results in a loss of the normal lactobacillus-dominantvaginal ¯ora, an elevated vaginal pH, increased introital colo-nization with E. coli, and increased UTI. Topical estrogen can
reverse these changes and greatly reduce the incidence of re-current UTI in postmenopausal women not using hormonereplacement therapy [17].
Complicated UTI
Clinical and epidemiologic studies of complicated UTIs have
been hindered by the heterogeneous nature of these conditions.
UTIs in diabetics, for example, are quite different than catheter-associated UTIs or UTIs in patients with spinal cord injury. Aworkable classi®cation of complicated UTIs is needed to fa-cilitate research in this area. Since most of these infections arenot seen in large numbers by any one institution, multicenterstudies will be needed. Studies of natural history and patho-genesis, as well as randomized clinical trials addressing man-agement, are needed. Infections in diabetics are a particularlyimportant group to study.
Catheter-Associated UTI
Studies over the last decade have stressed the importance of
bio®lm formation in the pathogenesis of catheter-associated
UTIs [18]. Bio®lms consist of Tamm-Horsfall protein, struviteand apatite crystals, bacterial polysaccharides and glycocalyces,
Downloaded from https://academic.oup.com/jid/article-abstract/183/Supplement_1/S1/2190986
by guest
on 20 March 2018

S4 Stamm and Norrby JID 2001;183 (Suppl 1)
and living bacteria. Microbes embedded in the bio®lm itself are
less susceptible to antimicrobials than planktonic bacteria andmay thus serve as a source of persistence leading to recurrenceafter antimicrobial therapy. Removal and replacement of acatheter at the time of treatment has been recommended bysome as an approach to dealing with this reservoir. Cathetermaterials that resist adhering bacteria or catheters that incor-porate antimicrobials or silver ions have also been developedto prevent catheter-associated UTIs. It has recently becomeapparent that organisms in the bio®lms communicate with oneanother via quorum sensing. Interference with these molecularsignaling mechanisms may be another way to disrupt bio®lmsand prevent related infections.
Prevention
Despite considerable advances in our understanding of the
pathogenesis of UTI, we have not yet developed new, clinically
useful means for preventing UTIs. Low-dose antimicrobial pro-phylaxis remains effective in many clinical settings and can besafely used for years in patients with recurrent UTIs; however,increasing antimicrobial resistance may eventually limit its ef-®cacy. Behavioral interventions (e.g., cessation of spermicideor diaphragm/spermicide use or reduction in coital frequency)may be effective but have not been actually studied as inter-ventions. In postmenopausal women, topically applied intra-vaginal estrogen has been effective in substantially reducing theincidence of recurrent UTI. Cranberry juice has been advocatedas a foodstuff that may prevent UTI by serving as a competitiveinhibitor of bacterial attachment to epithelial cells. One con-trolled trial suggests that regular ingestion of cranberry juicein older women with asymptomatic bacteriuria can reduce theprevalence of asymptomatic bacteriuria, but whether sympto-matic infections can be prevented has not been determined [19].
Another approach to prevention that is being actively pur-
sued is the use of a lactobacillus probiotic, namely oral orvaginal application of speci®cally selected lactobacillus strainswith the intent of restoring the normal, lactobacillus-dominant¯ora [20]. Lactobacilli utilized as probiotics must adhere wellto vaginal cells and have antimicrobial mechanisms, such assurfactant and H
2O2production, which may prevent E. coli
colonization and UTI. This approach, as well as the use ofcarefully selected avirulent E. coli strains to colonize the bladder
and prevent colonization with virulent strains, are discussedelsewhere in this supplement. With progressive clari®cation ofthe roles of speci®c adhesins in the pathogenesis of UTI, anti-adhesin vaccines are being developed to prevent UTIs inwomen. For Type 1 pilus±mediated attachment, there is littlestrain-to-strain variation in the adhesin protein itself, and, thus,induction of immunity against multiple strains would be fea-sible and is now being pursued. New approaches to preventions,such as those outlined above, are needed to signi®cantly impactthe continued high incidence of UTI.
References
1. Harding GKM, Ronald AR. The management of urinary infections: what
have we learned in the past decade. Int J Antimicrob Agents 1994;4:83±8.
2. Kunin CM. Urinary tract infections. Baltimore: Williams & Wilkens, 1997.
3. Schappert SM. National ambulatory medical care survey: 1992 survey.
Hyattsville, MD: National Center for Health Statistics, 1994 ; DHHS pub-
lication no. (PHS) 94±1250. (Advanced data for vital and health statistics.
No. 253).
4. Hooton TM, Scholes D, Hughes JP, et al. A prospective study of risk factors
for symptomatic urinary tract infection in young women. N Engl J Med1996;335:468±74.
5. Foxman B, Frerichs RR. Epidemiology of urinary tract infection. 1. Dia-
phragm use and sexual intercourse. Am J Public Health 1985;75:1308±13.
6. Krieger JN, Ross SO, Simonsen JM. Urinary tract infection in healthy uni-
versity men. J Urol 1993;149:1046±8.
7. Stamm WE, Hooton TM. Management of urinary tract infections in adults.
N Engl J Med 1993;329:1328±34.
8. Stamm WE. Urinary tract infections. In: Root RK, Waldvogel F, Corey L,
Stamm WE, eds. Clinical infectious diseases: a practical approach. NewY ork: Oxford University Press, 1999:649±56.
9. Saint S, Scholes D, Fihn S, Farrell R, Stamm W . The effectiveness of a clinical
practice guideline for the management of presumed uncomplicated urinarytract infection in women. Am J Med 1999;106:636±41.
10. Wong ES, McKevitt M, Running K, et al. Management of recurrent urinary
tract infections with patient-administered single-dose therapy. Ann InternMed 1985;102:302±7.
11. Schaeffer AJ, Stuppy BA. Ef®cacy and safety of self-start therapy in women
with recurrent urinary tract infections. J Urol 1999;161:207±11.
12. Warren JW , Abrutyn E, Hebel JR, Johnson JR, Schaeffer AJ, Stamm WE.
Guidelines for antimicrobial treatment of uncomplicated acute bacterialcystitis and acute pyelonephritis in women. Clin Infect Dis 1999;29:
745±58.
13. Gupta K, Scholes D, Stamm WE. Increasing prevalence of antimicrobial
resistance among uropathogens causing acute uncomplicated cystitis inwomen. JAMA 1999;281:736±8.
14. Talan, DA, Stamm WE, Hooton TM, et al. Comparison of cipro¯oxacin (7
days) and trimethoprim-sulfamethoxazole (14 days) for acute uncompli-cated pyelonephritis in womenÐa randomized trial. JAMA 2000;283:
1583±90.
15. Echols RM, Tosiello RL, Haverstock DC, Rice AD. Demographic, clinical
and treatment parameters in¯uencing the outcome of acute cystitis. ClinInfect Dis 1999;29:113±9.
16. Stamm WE, Raz R. Factors contributing to susceptibility of postmenopausal
women to recurrent urinary tract infections. Clin Infect Dis 1999;28:
723±5.
17. Raz R, Stamm WE. A controlled trial of intravaginal estriol in postmeno-
pausal women with recurrent urinary tract infections. N Engl J Med1993;329:753±6.
18. Stamm WE. Urinary tract infections. In: Bennett JV, Brachmann PS, eds.
Hospital infections. 4th ed. Philadelphia: Lippincott-Raven, 1997:477±85.
19. Avorn J, Monane M, Gurwitz JH, et al. Reduction of bacteriuria and pyuria
after ingestion of cranberry juice. JAMA 1994;271:751±4.
20. Redondo-Lopez V, Cook RL, Sobel JD. Emerging role of lactobacilli in the
control and maintenance of the vaginal bacterial micro¯ora. Rev InfectDis1990;12:856±72.
Downloaded from https://academic.oup.com/jid/article-abstract/183/Supplement_1/S1/2190986
by guest
on 20 March 2018