Bacterial Interference for Prevention of UTIs
Discussion
Bacterial interference has been touted as a promising approach to UTI prevention for the future (Biering-Sorensen 2002). In this approach, a group of bacteria that do not cause UTIs are introduced into the bladder which acts to limit the ability of other pathogens to effectively colonize the bladder and cause a symptomatic UTI. To date, the specific approach employed in studies in persons with SCI has been to colonize the bladder with either E. coli 83972 (Hull et al. 2000; Darouiche et al. 2005; Prasad et al. 2009) or E. coli HU2117 (Darouiche et al. 2011; Traunter et al. 2007). Most notably, Darouiche et al. (2005) conducted a prospective, randomized, placebo-controlled, double-blind trial (n=27) in which they randomized persons with SCI of greater than 1 years duration and with a history of symptomatic UTIs to receive bladder inoculation of either E. coli 83972 or sterile normal saline at a 3:1 ratio. This was preceded by a one-week course of empirically selected antibiotics as it had been noted that successful colonization is more likely achieved with a sterile bladder (Hull et al. 2000). Patients were monitored over the following year with monthly urine cultures. The number of UTIs experienced by those with successful E. coli 83972 colonization had significantly fewer UTIs than those with saline inoculation or unsuccessful E. coli inoculation (1.6 versus 3.5 episodes/year, p=0.036). The period during which the bladder remained colonized by E. coli 83972 was variable among study participants with only 13 of 21 patients being successfully colonized for at least 1 month, 4 of these remaining colonized for the entire 1 year study period and 9 losing E. coli after an average of 3.5 months. It should be noted that statistical comparisons were made between those with successful colonization (n=13) versus those inoculated with saline (n=6) combined with those not successfully inoculated (n=8). Only 1 of the 13 participants successfully inoculated developed a UTI while E. coli 83972 was in the bladder and this was associated with another organism (P. aeruginosa). No adverse events were obtained with the E. coli 83972 inoculations although 1 person in the saline group developed autonomic dysreflexia which subsided post-inoculation. Using a less robust pre-post study design, Prasad et al. (2009) also reported that preinoculation antibiotics improved inoculation rates and that rates of UTI declined during the period of colonization.
A longer period of colonization was achieved in a pre-post study conducted by Hull et al. (2000) in which 21 individuals with longstanding SCI (>18 months) and a history of symptomatic UTI over the preceding year were inoculated with E. coli 83972 following a course of appropriate antibiotics for 5-7 days. Persistent colonization of greater than 1 month was achieved in 13 study participants with mean colonization duration of 12.3 months (range 2-40 months). No participant sustained a UTI while colonized with E. coli even though these same individuals had a mean of 3.1 UTIs over the previous year. UTIs were noted in 4 of 7 persons not successfully colonized and at a rate of 3.5 UTIs/year for the months following loss of colonization in those where E. coli 83972 was no longer found in the bladder. The overall results from these three studies point to a strong effectiveness associated with this approach while the bladder remains colonized but that more work is required to enhance the rate of successful inoculation and to examine methods for sustaining the period of colonization.
An additional strain, E.coli HU2117, when colonized has demonstrated efficacy in preventing symptomatic UTIs from other uropathogens. Traunter et al. (2007) reported that in a group of adult SCI patients, who had been injured for one year or longer, colonization of E.coli HU2117 lead to decreases in the uropathogen Proteus. Further, a RCT conducted by Darouiche et al. (2011) reported that mean number of UTIs per patient was significantly less for those treated with E.coli HU2117 bacterial interference compared to individuals receiving a placebo inoculation.
Conclusion
There is level 1b evidence (from one RCT and two pre-post studies: Darouiche et al. 2005; Hull et al. 2000; Prasad et al. 2009) that bacterial interference in the form of E. coli 83972 bladder inoculation may prevent UTIs.
There is level 1b evidence (from one RCT and one pre-post: Darouiche et al. 2011; Trautner et al. 2007) that bacterial interference in the form of E.coli HU2117 bladder inoculation may prevent UTIs.
