Recurrent urinary tract infection incidence rates decrease in women with cystitis cystica after treatment with D-mannose: A cohort study

Abstract

Objectives:

D-mannose is a promising non-antibiotic prophylaxis for recurrent urinary tract infection (rUTI). rUTI is common in postmenopausal women and may be especially prevalent in those with cystitis cystica (CC) lesions found on cystoscopy. Our objectives were to determine whether CC lesions are associated with a higher UTI incidence rate and whether D-mannose reduces this rate in women with CC.

Methods:

This is a retrospective cohort study of rUTI patients who underwent cystoscopy at our institution (from which CC status was identified) and who were treated with D-mannose as a single agent for UTI prophylaxis. Participants were required to have at least one year of follow-up for UTIs both before and after D-mannose initiation to allow for a pre-post comparison.

Results:

Twenty-seven patients were included in the analysis (13 with CC, 14 without CC). The majority (88.9%) of patients were postmenopausal. Patients with CC had a higher UTI incidence rate than patients without CC (4.69 vs. 2.93 UTIs/year before starting D-mannose prophylaxis, P = 0.021). After initiating D-mannose prophylaxis, the UTI incidence rate decreased significantly in patients with CC (rate decrease=2.23 UTIs/year, P = 0.0028). This decrease was similar in magnitude to that observed in patients without CC (rate decrease=1.64 UTIs/year, P = 0.0007; P-interaction = 0.58).

Conclusions:

Recurrent UTI patients with CC had more frequent UTI episodes than patients without CC. Patients in both groups had fewer UTI episodes after beginning D-mannose prophylaxis. These findings add to the body of literature supporting D-mannose for prevention of recurrent UTI in women, including those with CC.

Introduction:

Urinary tract infections (UTIs) are the most common infection among women. Eleven percent of women in the United States have at least one physician-diagnosed UTI per year, and 60% of women will have a UTI at some point during their lifetime¹. In addition to costing an estimated $2.4 billion or more annually to treat², UTIs can negatively affect women’s quality of life, ability to attend to work and personal responsibilities, and sexual function^3,4. Moreover, a large analysis of MarketScan data from 2004 through 2010 found that 102 per 100,000 women had recurrent UTI (rUTI), defined as two episodes of acute cystitis within six months or three episodes within one year⁵. rUTI are especially prevalent in postmenopausal women, reaching 189 per 100,000 women aged 55–64 in the MarketScan analysis⁵.

In our clinical practice, we have often observed cystitis cystica (CC) on cystoscopy in postmenopausal women with rUTI. CC (known pathologically as follicular cystitis) appears as nodules in the bladder epithelium on cystoscopy. CC nodules contain lymphoid cells^6,7 and have often been considered as variations of normal, pre-malignant lesions, or non-specific responses to chronic inflammation^8-12. In pediatric patients, the number of bladder nodules seen on cystoscopy has been found to correlate with the number of UTI episodes in the preceding year⁷, possibly suggesting that bladder nodules increase the risk of UTIs in younger patients. However, no studies, to our knowledge, have investigated the relationship between bladder nodules or CC and UTI incidence in adult women.

Another unaddressed question is the best way in which to prevent rUTI in women with CC. The mainstay of acute UTI treatment and rUTI prevention in women is antibiotic therapy. However, antibiotic resistance is rapidly rising among common uropathogens¹³, and long-term antibiotic use is associated with serious side effects^14,15. Thus, many clinicians and researchers are working to identify and test non-antibiotic treatments to prevent rUTI. One non-antibiotic prophylaxis that has shown some promise in preventing rUTI is the natural sugar D-mannose, which is available as an over-the-counter dietary supplement^16-21. A systematic review and meta-analysis confirmed the ability of D-mannose to prevent rUTI in mainly pre-menopausal women unselected for or with unknown CC status²¹.

Although no studies, to our knowledge, have investigated the efficacy of D-mannose in women with CC, this non-antibiotic therapy may be particularly useful for women with CC for two reasons. First, by binding to the adhesin protein FimH on type 1 fimbriae, D-mannose may prevent common uropathogens such as Escherichia coli from adhering to bladder epithelial cells and establishing a UTI²². Second, D-mannose may also function by reducing inflammation, thereby possibly improving CC lesions and reducing the occurrence of UTIs (if CC lesions do, in fact, increase the risk of UTIs). This mechanism is supported by both in vitro and mouse model data. Zhang et al. reported that D-mannose stimulated differentiation of anti-inflammatory regulatory T cells in vitro and suppressed autoimmune diabetes and airway inflammation in mice²³. Likewise, D-mannose decreased autoimmune phenotypes in a mouse model of lupus²⁴. Finally, D-mannose reduced secretion of pro-inflammatory cytokines in macrophages²⁵.

Given the paucity of data on CC and D-mannose therapy, we took advantage of existing data in the Women’s Genito-Urinary Tract Specimen Consortium (WGUTSC) biobank to begin to investigate: 1) the relationship between CC and UTI incidence in women with rUTI, and 2) the effect of D-mannose therapy on UTI incidence in these women.

Materials and Methods:

We conducted a retrospective cohort analysis within the WGUTSC biobank at Washington University School of Medicine between 2015 and 2018. All patients attending the Female Pelvic Medicine and Reconstructive Surgery division clinic were invited to contribute specimens to WGUTSC. Women who consented provided serum/plasma, peripheral blood mononuclear cells, and urine specimens, as well as access to their electronic medical record for research purposes. Upon enrollment, basic demographic and clinical information including medical comorbidities, past surgical history, past cystoscopy findings, and past UTI history were collected from the medical record. To date, 264 participants have enrolled in WGUTSC. For this study, we included participants if they had a diagnosis of rUTI (n=192), had a cystoscopy performed for history of refractory rUTI by our division between 2015 and 2018 (n=146), were taking D-mannose for UTI prevention based on documentation in the medical record and had at least one year of follow-up both before and after initiating D-mannose to allow for a pre-post comparison (n=27). We excluded patients who were also taking vaginal estrogen or methenamine. Washington University in St. Louis Institutional Review Board approval was obtained before data collection (#201712113, #20190534).

For the patients who met criteria for this retrospectively assembled cohort (n=27), we identified presence or absence of CC on cystoscopy report. CC diagnoses were made by the provider (one of three Female Pelvic Medicine and Reconstructive Surgeons) at time of cystoscopy following a standard definition: the presence of translucent submucosal cysts that may form a cobblestone pattern or polypoid mass. The WGUTSC database and electronic medical record were used to collect date of cystoscopy, date of D-mannose initiation, and number of UTI episodes (defined as >10⁵ colony-forming units/mL of urine) documented in the 12 months before and after starting D-mannose. Patients were given a standing urine culture order to leave a urine specimen when they experienced symptoms, and these results were reported in their medical record. If they used a laboratory that was not affiliated with our institution or were treated without a proven culture, this information could not be ascertained from the medical record and was not included in our analysis.

Fisher’s exact tests, Student’s t-tests, and Mann-Whitney U tests were performed to compare presence or absence of CC with clinical and demographic variables. Incidence rates were calculated and compared by Poisson regression with robust variance estimation to account for small deviations from a Poisson distribution and repeated measures. Statistical significance was set at P < 0.05. All analyses were conducted in SAS version 9.4 (SAS Institute Inc., Cary NC, USA. Copyright 2016).

Results:

Twenty-seven participants in the WGUTSC database met inclusion criteria for this retrospective cohort analysis. Thirteen had evidence of CC on cystoscopy, and fourteen had no evidence of CC on cystoscopy. Nineteen had cystoscopy performed prior to D-mannose initiation, seven after D-mannose initiation, and one on the same day as D-mannose initiation. Patients with and without CC were overall similar in demographic characteristics and past medical, surgical, and social history (Table 1). The median age of our patients was 73 years. Most patients were white (88.9%), postmenopausal (88.9%), and nonsmokers (88.9%). All patients (100%) with CC and 78.6% of patients without CC were postmenopausal, though this difference was not significant (P = 0.22). Patients with CC were more likely than those without CC to have a history of urinary incontinence (46.2% vs 7.1%, P = 0.033) and adnexal surgery (53.9% vs 14.3%, P = 0.046).

Table 1:

Demographic, social, and clinical characteristics of female recurrent urinary tract infection patients with and without CC, Women’s Genito-Urinary Tract Specimen Consortium biobank

		Total (n=27)	With CC (n=13, 48.2%)	Without CC (n=14, 51.9%)	P-value
Age, median (interquartile range)		73 (67-79)	73 (70-77)	76 (60-82)	0.36
Race	White	24 (88.9)	11 (84.6)	13 (92.9)	0.60
	Unknown	3 (11.1)	2 (15.4)	1 (7.1)
Postmenopausal	Yes	24 (88.9)	13 (100)	11 (78.6)	0.22
	No	3 (11.1)	0 (0.0)	3 (21.4)
Tobacco use	Never smoker	24 (88.9)	12 (92.3)	12 (85.7)	1.00
	Former smoker	3 (11.1)	1 (7.7)	2 (14.3)
Body mass index (kg/m2), mean		27.5	29.6	25.6	0.065
Medical history	Hypertension	8 (29.6)	6 (46.2)	2 (14.3)	0.10
	Hyperlipidemia	9 (33.3)	5 (38.5)	4 (28.6)	0.69
	Diabetes	4 (14.8)	3 (23.1)	1 (7.1)	0.33
	Hypothyroidism	6 (22.2)	2 (15.4)	4 (28.6)	0.65
	Chronic pelvic pain	5 (18.5)	2 (15.4)	3 (21.4)	1.00
	Pelvic organ prolapse	7 (25.9)	4 (30.8)	3 ((21.4)	0.68
	Urinary incontinence	7 (25.9)	6 (46.2)	1 (7.1)	0.033
Surgical history	Hysterectomy	14 (51.8)	9 (69.2)	5 (35.7)	0.13
	Adnexal surgery	9 (33.3)	7 (53.9)	2 (14.3)	0.046
	Vaginal prolapse repair	5 (18.5)	4 (30.8)	1 (7.1)	0.16
	Urinary incontinence procedure	8 (29.6)	6 (46.2)	2 (14.3)	0.10

Except where noted, all data are presented as n (%). CC, cystitis cystica.

Patients with CC had a higher UTI incidence rate than patients without CC (4.69 vs. 2.93 UTIs/year before starting D-mannose prophylaxis, P = 0.021, Table 2). The UTI incidence rate decreased significantly after initiating D-mannose prophylaxis in women with CC (rate decrease=2.23 UTIs/year, P = 0.0028). This decrease was similar in magnitude to that observed in women without CC (rate decrease=1.64 UTIs/year, P = 0.0007; P-interaction = 0.58).

Table 2:

UTI incidence rates before and after D-mannose treatment in recurrent urinary tract patients with and without CC, Women’s Genito-Urinary Tract Specimen Consortium biobank

	UTI incidence rate/year
	Before D-mannose	After D-mannose	Change in UTI incidence rate/year	P-value*
With CC (n=13)	4.69	2.46	2.23	0.0028
Without CC (n=14)	2.93	1.29	1.64	0.0007

^*Calculated by Poisson regression with robust variance estimation. CC, cystitis cystica. UTI, urinary tract infection

Discussion:

In our study, rUTI patients with CC had more frequent UTIs than patients without CC. We also found that D-mannose reduced UTI incidence rates for patients with CC, as well as for those without CC.

Our study provides evidence that CC may be associated with more frequent UTIs in women with rUTI. However, because CC was ascertained after the onset of rUTI, we cannot determine whether CC increases the incidence rate of UTIs or whether greater UTI incidence rates contribute to CC – or stated differently, whether CC is a risk factor for or consequence of greater UTI rates. Additional studies will be necessary to distinguish these two possibilities. However, data from older literature and mouse model studies support the plausibility of CC as a risk factor for UTIs. Currently, CC is often considered a benign bladder lesion of no clinical significance and not requiring treatment^9,10, but older literature indicates that these lesions are lymphoid tissue^6,12. Furthermore, Ligon et al. found that aged mouse bladders harbor organized bladder tertiary lymphoid tissues (bTLTs), and that aged mice have higher rates of rUTI than young mice when experimentally infected¹¹. If future clinical and translational research in humans supports this conclusion and reveals that CC lesions are, in fact, bTLTs and risk factors for UTIs, then clinicians may need to consider performing cystoscopy in postmenopausal women with rUTI. Currently, only 2.8% of women with rUTI undergo cystoscopy⁵, but this low rate of cystoscopy may underdiagnose a higher CC prevalence.

A second major finding of our study is that women with and without CC had fewer UTI episodes after initiating D-mannose than before. Although this finding could potentially be explained by confounding by other preventive factors initiated at the same time as D-mannose, the fact that we limited the analysis to women not taking any other preventive medications (i.e., vaginal estrogen or methenamine) makes this possibility less likely. Our finding is also in line with clinical trial reports supporting the clinical efficacy of D-mannose for preventing UTI, particularly in women using D-mannose as their sole prophylactic therapy^18-21. In our study, we also found that D-mannose use was associated with fewer UTI episodes in women with CC, supporting its use in this previously un-studied group of women.

One strength of our study is that many patients were postmenopausal, the life stage at which women are most at risk for rUTI. Additionally, we used strict criteria for documenting UTI episodes with urine culture-proven infections (>10⁵ colony-forming units/mL) and excluded patients on other medications for UTI prophylaxis. We did so because therapies such as vaginal estrogen also affect the inflammatory profile²⁶ and may have a synergistic effect with D-mannose. However, a resulting limitation of our study was the small sample size, as few patients use D-mannose as a single agent therapy for UTI prevention. In our practice, patients diagnosed with rUTI are recommended vaginal estrogen as first-line prophylaxis. We often advise patients to take D-mannose at the same time as we prescribe vaginal estrogen or add D-mannose if vaginal estrogen alone does not prevent rUTI. However, patients who decline or are unable to use vaginal estrogen (i.e., history of estrogen-dependent malignancy) are candidates for D-mannose. Our study also does not account for possible behavioral modifications that patients may have made for UTI prevention (i.e., voiding after intercourse). Another limitation of our study was that we could not control the dose of D-mannose that patients took, as they purchased it as an over-the-counter supplement. Our printed instructions inform patients that the current recommended dose is one gram twice daily or two grams once daily; however, the actual dosage used was not documented in the electronic medical record. In the seven patients with cystoscopy performed after starting D-mannose, there is also a theoretical possibility that D-mannose could affect development or resolution of CC. However, this has not been studied in the current literature, and future studies should explore this potential relationship. Lastly, documented UTI episodes were limited to cultures performed at laboratories associated with our institution. The number of UTI episodes that may not be accounted for is unknown; however this likely applies to both patients with and without CC.

In conclusion, our observational data suggest that CC may be related to UTI frequency, and that D-mannose appears to work as well in rUTI women with CC as in those without CC. Future prospective observational studies and/or randomized controlled trials could be conducted to confirm these conclusions. Moreover, translational studies of CC lesions could be conducted to determine the mechanisms of CC and D-mannose activity. These findings add to the body of literature supporting D-mannose for prevention of rUTI in women, including those with CC.