Reinfection versus relapse in urinary tract infection.

Sir—The much-needed trial by Dow et al. [1] regarding treatment duration for urinary tract infection (UTI) in patients with spinal cord injury and the accompanying editorial commentary [2] raise questions regarding the determination of whether isolates from recurrent (posttherapy) episodes of UTI represent relapse or reinfection and determination of the possible sources of such infections. First, although various methods can be used to classify isolates from recurrent episodes of UTI as the same strain or a different strain than the pretherapy isolate, even isolates of the same strain can represent reinfection if there is a persisting external reservoir from which the organism can be reintroduced into the host’s urinary tract. Thus, whereas isolates of different strains almost certainly represent reinfection, isolates of the same strain are ambiguous with respect to whether they represent reinfection or relapse. This ambiguity can lead to overestimation of relapse rates. Second, because of the relatively high prevalence of Klebsiella species, Enterococcus species, and Escherichia coli among pretherapy isolates from urine in the study of Dow et al. [1], an inference that a posttherapy isolate of the same species represents the same strain as the pretherapy isolate may be erroneous. Subspecies typing methods, such as PFGE, are needed here. It is unclear why Dow et al. [1] reserved PFGE analysis for only selected same-species posttherapy isolates. Third, even PFGE analysis may not provide unambiguous results. For example, it is statistically improbable that a patient whose pretherapy urine sample yielded Acinetobacter anitratus would have recurrent UTI due to an unrelated strain of A. anitratus, given that the overall prevalence of infection by Acinetobacter species in the study population was 10%, and only a subset of those infections, presumably, was due to A. anitratus [1]. It is perhaps as likely that the PFGE results in this instance were in error, or that the strain, while residing in the patient (or in a patient-associated reservoir), underwent genetic rearrangements that produced the observed PFGE profile alterations, which led to a false assessment that this was a different strain from the pretherapy isolate. Finally, 70% of the study subjects were men [1]. Because the prostate gland is a common source for relapsing UTI in men [3] and is usually involved in cases of febrile UTI in men, despite the absence of localizing symptoms [4], it may be that some of the relapses observed by Dow et al. [1] derived from a persistently infected prostate rather than the upper urinary tract, as was proposed. The tenacity of prostatic infection [3] would be consistent with the superior microbiological efficacy of the longer treatment course that was observed by Dow et al. [1].