‘Urinary Tract Infection’ and the Microbiome
The current paradigm for managing uncomplicated “urinary tract infection” (“UTI”) is deeply flawed. “UTI” is ambiguously defined, and coupled with a belief that “bacteria are not normal inhabitants of the urinary tract,”1 the diagnosis often leads to unnecessary, harmful antibiotic treatment. Although bacteriuria identified by standard clinical cultures (which we will call standard bacteriuria) is central to most definitions, more sensitive diagnostic tests now demonstrate that “urine is not sterile”2 and that standard bacteriuria represents a fraction of the diverse microbiota hosted by the urinary tract. Knowledge of this complex, generally beneficial microbiome deeply undermines the current paradigm, which relies on the findings of standard culture. By acknowledging this microbiome, a successor paradigm will generate new questions about relationships among host, microbiome, and antibiotic use, and will almost surely show additional serious harms from antibiotic overtreatment.
Shifting the paradigm for managing the urinary microbiome will be slow and difficult. Emphasizing the current paradigm’s shortcomings may reduce antibiotic overuse in the short run, and in the long run help in development of a successor.
This discussion concerns medically stable, nonpregnant adults with normal urinary tract structure and function. The role of antibiotics in patients with abnormalities of anatomy or physiology, such as spinal cord injury, urinary obstruction, or catheters, will require careful investigation. New insight into pyelonephritis and bacteremic bacteriuria is likely to develop.
Diagnosis and Management of “UTI”
The ambiguous definition of “UTI” seems to promote antibiotic overuse. In one common usage, “urinary tract infection is defined as microbial infiltration of the normally sterile urinary tract.”3 With this definition, asymptomatic bacteriuria is a “UTI” and is often treated, even in patient groups where strong evidence shows lack of benefit.4 A second common definition, “significant bacteriuria in a patient with symptoms or signs attributable to the urinary tract and no alternate source”1 seems more restrictive but does not define what symptoms or signs may be attributed to the urinary tract. This ambiguity creates opportunities for overtreatment. “UTI” may be suspected in patients without urinary tract symptoms who develop any of a wide variety of nonspecific findings, such as change in cognitive or physical function, or change in the appearance or odor of urine. The diagnosis is confirmed (using the first definition) if bacteriuria is present. To satisfy the second definition a clinician need only attribute the nonspecific findings to the urinary tract. Antibiotic treatment of “UTI” often follows, even though no data have shown that these changes respond to treatment.
Canonically, “all symptomatic UTI should be treated,”5 but actual benefit is limited. Hooton6 emphasizes that in acute uncomplicated cystitis, “the primary goal of treatment is to ameliorate symptoms.” Foxman7 summarizes that symptoms are usually self-limited, of brief duration, and only slightly shortened by antibiotic treatment; that cystitis rarely progresses to pyelonephritis; and that randomized trials show no reduction in the risk of progression to pyelonephritis with antibiotic treatment. The generally benign (other than symptoms) nature of “symptomatic UTI” is suggested by the billions of persons around the world and over the eons who have suffered “UTI” without access to antibiotics and have recovered fully.
With its various meanings, convenient diagnosis, long tradition, suggestive link to treatment, and uncritical acceptance by clinicians, patients, families, and insurers, “UTI” remains heavily embedded in practice; “one of the most common bacterial infections worldwide.”3 The paradigm provides tidy management for a patient with “UTI” who expects antibiotics. Further, the current paradigm does account for several findings. Standard bacteriuria is associated with pyuria, fever, and dysuria, for example; and these often improve with treatment, as do a wide variety of findings seemingly unconnected with the urinary tract. Antibiotic treatment improves outcomes for asymptomatic pregnant women who have standard bacteriuria. Pyelonephritis and bacteremic bacteriuria probably arise in the urinary tract and do require antibiotic treatment.
To diagnose “UTI” and determine antibiotic sensitivity based on results of standard cultures, however, is to rely on familiar, accessible data and to ignore the dozens of bacterial species,2 as well as intracellular bacterial colonies and urinary virome5 known to reside in the urinary tract. Current discussions of symptomatic or asymptomatic bacteriuria or sterile urine are similarly problematic. To attribute delirium to standard bacteriuria seems unjustifiable, knowing that most or all people with or without delirium have bacteriuria. The current paradigm is defensible only if all pathogenic organisms are identified with standard cultures and all organisms more difficult to identify can be safely ignored.
We propose instead that urinary symptoms, bacteremia, pyelonephritis, and other recognizable disturbances of the urinary tract are the dysbiotic tip of a much larger iceberg of complex host–microbe interactions that are occurring out of sight of standard cultures. As expected in the era of the microbiome, stable bacterial communities are generally beneficial. For example, compared with the instillation of sterile saline, “bladder colonization with (the nonpathogenic) E. coli HU2117 safely reduces the risk of symptomatic urinary tract infection in patients with spinal cord injury.”8 Of 699 young women with asymptomatic bacteriuria, half of whom were randomized to receive no antibiotic treatment, “treatment was associated with a higher rate of symptomatic UTI … (thus) asymptomatic bacteriuria … may play a protective role in preventing symptomatic recurrence” during 12-month follow-up.9
The normal urinary tract is an open system bearing urine from renal tubules to urethra without discrete anatomic interruption. Sensitive diagnostic techniques, for example, gene-sequencing and expanded quantitative urine cultures, confirm that the tract hosts diverse bacterial communities.2 The respiratory tract–open from alveolus to mouth, traditionally considered sterile, now known to host complex microbial communities–may provide useful analogies for investigation. Could viruses, common respiratory pathogens, be pathogenic in the urinary tract? Might cystitis be analogous to pharyngitis? Might uropathogenic E. coli be analogous to pneumococcus, prevalent and sporadically pathogenic? Do fever, pyuria, duration of symptoms or other clinical features help distinguish between patients who would do well with urinary tract analgesia and those who would be safer with antibiotic treatment?
To read this article in its entirety please visit our website.
-Thomas E. Finucane, MD
This article originally appeared in the March 2017 issue of The American Journal of Medicine.