Differentiating relapse from same-strain reinfection in recurring gram-negative bacteremia.

SIR—Pulsed-field gel electrophoresis of macrorestricted genomic DNA, the molecular strain typing method used by Wendt et al. [1, 2] to classify isolates recovered during recurrent bacteremia as “same” versus “different” in comparison with a patient’s previous blood isolate, is a powerful tool that is well suited to this purpose. However, no molecular typing method can define the mechanism responsible for the second appearance of a given strain in an individual patient, which instead becomes a matter of judgment that should be guided by epidemiological data and surveillance cultures. Although same-strain recurrences can be due to persistence of the organism within the host (i.e., a true “relapse” as assumed by Wendt et al. [1, 2]), reintroduction of the same strain from a persisting external reservoir, whether in the environment or in the patient’s own bacterial flora, is also possible. This latter scenario constitutes “same-strain reinfection” and is indistinguishable by molecular techniques from true relapse. Early same-strain reinfection is probably the most common cause of what masquerades as relapse in women who receive single-dose antimicrobial therapy for uncomplicated cystitis. In such cases, the initial urine pathogen commonly persists in the vagina and rectum, despite its disappearance from the urine, and can reenter the urinary tract once antimicrobial activity there has waned [3]. The distinction between true relapse and same-strain reinfection is of more than academic importance, since different interventions are required for each condition. Intensified therapy designed to eliminate a persisting internal focus of infection will be of little value if it fails to address a persisting external reservoir that is the true cause of same-strain recurrence. Infection control efforts may be highly relevant for same-strain recurrences that involve a persisting reservoir outside the host, whereas they would be irrelevant for true relapses from an internal focus. Wendt et al. [1, 2] noted that retained vascular catheters were a unifying theme among their subjects with presumed relapsing gram-negative bacteremia. This clinical circumstance indeed would argue for relapse rather than same-strain reinfection as the mechanism for the observed same-strain recurrences. However, it should be recognized that this probability remains an inference based on epidemiological data, not a direct conclusion from molecular strain typing.