Bacterial cystitis occurs with reasonable frequency in dogs. Persistent infection and repeated reinfection (recurrent bacterial urinary tract infection, hereafter termed recurrent UTI) is reported less frequently. Recurrent urinary tract infection is predominantly diagnosed in dogs with a predisposing or comorbid condition, and episodes are often subclinical. Escherichia coli is the most frequently cultured species in both uncomplicated bacterial cystitis and recurrent UTI. Antimicrobial therapy limitations and global emergence of bacterial strains that are both virulent and multidrug-resistant (MDR) has led to an urgent requirement for novel control strategies in both companion animals and humans. Escherichia coli strain 83972 has been successfully used to colonise the bladder of humans with recurrent UTI in order to reduce the incidence of subsequent febrile UTI via the process of active bacterial interference. This strategy has not been utilised in dogs to date but may prove an effective alternative to antimicrobial therapy if it can be shown to colonise the bladder of dogs safely and effectively compete with uropathogens.
A retrospective study to examine the signalment, infecting strains, presence of clinical signs, and incidence of pyuria in Australian dogs with recurrent urinary tract infection was performed. The findings were similar to those of two large North-American studies with regard to signalment, predisposing or comorbid diseases, bacterial species isolated, and resistance to commonly used oral antimicrobials, although the median age of our dogs was higher at 10 years (7 years in both prior studies). Examination of the E. coli isolates from a subset of nine dogs provided new information on phylogenetic group, virulence gene profiles, haemolytic phenotype, expression of type 1 fimbriae, and sequence type (ST). Phylogenetic groups F and B2 were frequently identified, representing 13/32 (41%) and 8/32 (25%) of isolates respectively, but no phylogroup B2 isolates were MDR, in comparison with phylogroup F isolates, which accounted for 12/16 (75%) MDR isolates. Clonal groups of global importance such as ST131-B2, ST354-F, ST38-D and ST457-F were represented, with possible implications for both companion animal and human health. Of particular concern was the repeated isolation of ST354-F strains from the urine of a paralysed dog over a 3-year period.
When competition experiments between E. coli 83972 and a selection of canine uropathogenic E. coli (UPEC), including emerging, highly virulent MDR strains with zoonotic potential, were performed in pooled canine urine, E. coli 83972 was largely non-competitive against the canine isolates after 17 hr of incubation. This finding contrasted with the competitive superiority previously exhibited by E. coli 83972 in pooled human urine in competition experiments with human UPEC strains. In 5/10 trials, the E. coli 83972 was not outcompeted in canine urine, however, and given that the starting ratio in dogs following prophylactic bladder colonisation would favour E. coli 83972, it is feasible that it could outcompete MDR canine UPEC strains in vivo.
A single dose inoculation trial of E. coli strain 83972 (2 X 108 viable cells) was performed in six healthy dogs. Although the duration of colonisation was short for the dogs as a group (0-10 days, median 4 days), two dogs remained colonised for at least 10 days. No dog suffered pyrexia or appeared systemically unwell but a small number displayed gross haematuria and/or stranguria. By day 3 of each trial (four dogs were re-inoculated on day 20) all clinical signs had resolved. As successful colonisation in humans is typically achieved using multi-dose protocols administered over several days, we subsequently aimed to design a protocol for canine inoculation that was likely to be more effective than a single inoculation but would be practical and acceptable to veterinarians and owners. Eight healthy dogs were inoculated with E. coli 83972 (1 X 109 viable cells) at 0, 8 and 24 hr via an indwelling urinary catheter. Although the majority of dogs quickly expelled the inoculated bacteria (median duration of colonisation 2 days), one dog was colonised for 28 days in the absence of any discernable anatomic abnormalities or other predisposing factors. Side effects encountered included pollakiuria or stranguria in 4/8 inoculated dogs and 1/3 sham-inoculated dogs but by day 8 of the trial all clinical signs had resolved.
In conclusion, the development of a non-antimicrobial strategy for the prevention of recurrent UTI in dogs is of importance to both canine and human health, and E. coli 83972 is a feasible candidate for use in the dog. In addition, further exploration of the ability of E. coli 83972 to compete with emerging clonal UPEC strains such as ST131 and ST1193 in both pooled human and canine urine may provide insight into why these emergent, multidrug-resistant high virulence-associated clonal lineages have become so successful in colonising and causing extraintestinal infection in individuals of either species.