To describe the urinary microbiota and to assess the performance of expanded quantitative urine culture (EQUC) on voided urines in women with recurrent urinary tract infection (RUTI).

This was an IRB-approved study of women ≥18 years old with RUTI (≥3 UTIs in prior 12 months or ≥2 UTI in prior 6 months). We excluded those with anatomic abnormalities of the urogenital tract, those with neurologic or immunologic disease, a history of bladder malignancy, or those with current systemic infection. In addition to demographic variables and current UTI self-report status, participants contributed voided and catheterized urine specimens. These specimens were submitted for culture by standard urine culture (SUC) as well as EQUC. Bacterial isolates were identified by MALDI-TOF mass spectrometry. Culture results were compared between SUC versus EQUC as well as catheterized versus voided urine for each participant. Data was analyzed by comparing species presence, abundance, and correlation with symptoms. Chi square and Fisher’s exact test were used to test for significance.

37 participants of average age of 67 years were included; most were Caucasian (78%), postmenopausal (84%), had taken antibiotics for a UTI within the last 30 days (76%), and were currently using vaginal estrogen (54%). Participants reported an average of 52 lifetime UTIs and an average RUTI duration of nine years. At the time of specimen collection, 13 women self-reported UTI, while 24 did not. Compared to SUC, EQUC reproducibly detected more bacterial species in both catheterized (44 vs. 15) and voided urine (156 vs. 47) samples. EQUC also detected a larger total number of potential uropathogens compared to SUC from both catheterized (25 vs. 13) and voided (71 vs. 26) specimens. From voided specimens, we cultured a larger total number of potential uropathogens than catheterized with both EQUC (71 vs. 25) and SUC (26 vs. 13); thus, we observed more total uropathogens in voided urine (71) than in catheterized urine (25) and detected some species in voided urine not seen in the catheterized specimens. Enterococcus faecalis was the most prevalent uropathogen and was detected by EQUC in 13 women, 6 times from catheterized urine and 13 times from voided urine. The detection of E. faecalis was related to self-report UTI status: 7/13 in positive self-report vs. 6/24 in negative self-report (p=0.08).  Escherichia coli, the second most prevalent uropathogen, was detected by EQUC in 11 women, 6 times from catheterized urine and 10 times from voided urine. Only 3/11 women self-reported UTI, and 2 of these women were also culture-positive for E. faecalis.

Compared to SUC, EQUC provides a more comprehensive description of the female urinary microbiota of RUTI patients. This is true for both catheterized and voided urine specimens and for both uropathogens and other bacterial species. Voided urine specimens contained the majority of uropathogens, raising the possibility that some uropathogens may not be inhabiting the bladder, but more distal regions of the urinary tract – such as the urethra – or possibly the vagina. In this cohort, E. coli was not the most frequently detected uropathogen and was not often associated with UTI symptoms, suggesting that RUTI may not be E. coli-centric.

The urinary microbiota of women with RUTI cannot be adequately assessed with SUC, especially when using voided urine.  Uropathogens were detected nearly three times as often in voided urine compared to catheterized urine, suggesting that diagnostic precision is enhanced with catheterized specimens. It also suggests that uropathogens that inhabit the urethra may cause or contribute to symptoms. Many women did not report UTI despite the presence of uropathogens, including the minority that was E. coli-dominant. These findings suggest that the disease mechanism of RUTI may be a polymicrobial process, with heightened dysbiosis associated with E. faecalis.