The Urinary Microbiome in Postmenopausal Women with Recurrent Urinary Tract Infections

Monique H. Vaughan, Jialiang Mao, Lisa A. Karstens, Li Ma, Cindy L. Amundsen, Kenneth E. Schmader, Nazema Y. SiddiquibioRxiv 2020 · DOI

TL;DR

This study investigated the urinary microbiome in postmenopausal women with recurrent UTIs, finding that while overall Lactobacillus abundance didn't differ significantly, specific anaerobic bacteria were associated with UTI status. These findings suggest that shifts in anaerobic bacteria, rather than just Lactobacillus levels, may play a crucial role in recurrent UTIs in this population.

What Problem Does This Paper Address?

Recurrent urinary tract infections (UTIs) are a significant problem for postmenopausal women, and their exact causes are not fully understood. While treatments like vaginal estrogen and prophylactic antibiotics are used, the role of the urinary microbiome (the community of microbes in the urinary tract) is increasingly suspected to be involved. Previous research has hinted at a decline in Lactobacillus species after menopause, which are generally considered beneficial. However, the specific microbial profiles associated with recurrent UTIs, especially in women undergoing treatment, remain unclear. This paper aims to compare the "steady state" urinary microbiome in three distinct groups of postmenopausal women: those with recurrent UTIs on daily antibiotic prophylaxis, those with recurrent UTIs not on antibiotics, and age-matched controls, all of whom were using vaginal estrogen.

Methods

The researchers employed a multi-pronged approach to characterize the urinary microbiome:

  • Expanded Quantitative Urine Culture (EQUC): This technique goes beyond standard urine cultures to detect a wider range of bacteria, including fastidious and anaerobic organisms, at both genus and species levels.
  • 16S rRNA Gene Sequencing: This culture-independent method provides a broader overview of the microbial community, allowing for the identification of a vast array of microbes, including those difficult to culture.
  • Bayesian Graphical Compositional Regression (BGCR): This advanced statistical method was used to analyze the microbiome data. BGCR is particularly useful for compositional data (where the relative abundance of microbes matters) and can identify specific taxa that differ between groups, while accounting for potential confounding factors.
  • Permutational Multivariate Analysis of Variance (PERMANOVA): This statistical test was used to compare the overall microbial communities between groups and to assess the influence of covariates like sexual activity, age, and BMI.

Datasets

The study utilized urine samples from 64 postmenopausal women. These women were categorized into three groups:

  1. Women with recurrent UTIs on daily antibiotic prophylaxis (rUTI+antibiotics, n=17).
  2. Women with recurrent UTIs not on antibiotic prophylaxis (rUTI, n=24).
  3. Age-matched controls without recurrent UTIs (controls, n=23).

All participants were using vaginal estrogen therapy. The data collected was cross-sectional, representing a baseline snapshot of the urinary microbiome.

Key Insights and Ideas

  • Beyond Lactobacillus Abundance: While previous research focused on the decline of Lactobacillus in postmenopause, this study suggests that the presence and abundance of specific anaerobic bacteria might be more critical in the context of recurrent UTIs, especially in women on antibiotic prophylaxis.
  • Antibiotic Prophylaxis's Impact: The study highlights that women on daily antibiotic prophylaxis for recurrent UTIs showed distinct differences in their anaerobic bacterial composition compared to controls, suggesting that long-term antibiotic use can significantly alter the urinary microbiome.
  • The Power of Advanced Analytics: The use of BGCR proved crucial in pinpointing specific anaerobic taxa (like Prevotellaceae, Clostridiales, and Actinobacteria) that were associated with different UTI phenotypes, offering a more nuanced understanding than traditional methods alone.
  • Concordance and Limitations of Methods: The comparison between EQUC and 16S rRNA sequencing revealed that while both methods identified some common microbes, 16S rRNA sequencing detected a broader range, including many unculturable organisms. However, certain bacteria, like some Enterococcaceae, were only identified by culture, underscoring the complementary nature of these techniques and the limitations of each.

Key Results

  • EQUC Findings: Using expanded quantitative urine culture, no significant differences were found in the median number of microbial species isolated among the three groups. Similarly, Lactobacillus abundance did not differ significantly between groups (p=0.72). However, there were trends suggesting differences in specific Lactobacillus species.
  • 16S rRNA Sequencing and BGCR: Bayesian graphical compositional regression analysis revealed significant differences in anaerobic taxa associated with the phenotypic groups. These differences were primarily centered on the order Bacteroidales and the family Prevotellaceae, with notable differences also observed in Actinobacteria and certain genera of Clostridiales.
  • Specific Lactobacillus Trends: While overall Lactobacillus abundance was similar, there were trends towards fewer L. gasseri and L. crispatus in the rUTI+antibiotics group compared to the other groups, though these did not reach statistical significance.
  • Covariate Influence: Sexual activity was significantly associated with increased differences in the urinary microbiome composition between groups. Age and BMI also approached statistical significance.

Limitations and Future Work

  • Small Sample Size: The authors acknowledge that the study's primary limitation is its small sample size, which may have prevented some observed trends (e.g., in Lactobacillus species) from reaching statistical significance. This also restricted the number of potential candidates due to strict entry criteria.
  • Cross-Sectional Design: The cross-sectional nature of the study means it cannot establish causality. It's unclear whether the observed differences in anaerobic bacteria are a cause or consequence of recurrent UTIs or antibiotic prophylaxis.
  • Standardization of Prophylaxis: The antimicrobial agents used for prophylaxis were not standardized, which could influence findings.
  • Future Directions: The study suggests that future research should focus on:
  • Investigating the complex interactions among bacteria, particularly the balance between Lactobacilli and anaerobic bacteria, in the context of recurrent UTIs.
  • Conducting longitudinal studies to track changes in the urobiome over time and establish causal relationships.
  • Exploring specific anaerobic taxa identified in this study for their potential role in UTI pathophysiology.
  • Employing additional methods, such as mechanical or enzymatic cell lysis, to better understand the Gram-positive component of the microbial community.

Summary updated

Abstract

ABSTRACT Recurrent urinary tract infections (UTI) are highly prevalent in postmenopausal women, where vaginal estrogen and prophylactic antibiotics are used for treatment. The etiology of recurrent UTIs is not completely known, but the urinary microbiome is thought to be implicated. Thus, we aimed to compare the “steady state” urinary microbiome in three groups of menopausal women who were all using topically-applied vaginal estrogen: 1) women with recurrent UTIs on daily antibiotic prophylaxis; 2) women with recurrent UTIs not on antibiotic prophylaxis; and 3) age-matched controls without recurrent UTIs. Here we present a cross-sectional analysis of baseline data from 64 women enrolled in a longitudinal cohort study. Catheterized urine samples were collected > 4 weeks after last treatment for UTI. Samples were evaluated using expanded quantitative urine culture (EQUC) and 16S rRNA gene sequencing. With EQUC techniques, there were no significant differences in the median numbers of microbial species isolated among groups (p=0.96), even when considering Lactobacilli (p=0.72). However, there were trends towards different Lactobacillus species between groups. With sequencing the overwhelming majority of urinary samples contained Lactobacilli , with non-significant trends in relative abundance of Lactobacilli among groups. Using a Bayesian regression analysis for compositional data, we identified significant differences in anaerobic taxa that were associated with phenotypic groups. Most of these differences centered on Bacteroidales and the family Prevotellaceae , though differences were also noted in Actinobacteria and certain genera of Clostridiales. Associations between anaerobes within the urinary microbiome and recurrent UTI warrants further investigation. IMPORTANCE In menopausal women with recurrent urinary tract infections (UTIs) compared to those without, the abundance of Lactobacillus within the urinary microbiome is not significantly different when vaginal estrogen is regularly used. In this population, Lactobacillaceae were identified in 97% of urine samples using culture-independent techniques. However, with expanded urine cultures, women with recurrent UTIs taking daily antibiotics had a disproportionately low amount of L. gasseri / L. acidophilus compared to the other phenotypic groups. These findings support the theory that certain Lactobacillus species may be more important than others in the pathophysiology of postmenopausal recurrent UTIs. Furthermore, when using culture-independent techniques to explore urinary microbiota across phenotypic groups, we identified differences in multiple anaerobic taxa. Taken together, these results suggest that altered ratios of anaerobes and certain Lactobacillus species within the urinary microbiome may be implicated in postmenopausal recurrent UTI.
BibTeX
@misc{vaughan2020urinary,
  author = {Monique H. Vaughan and Jialiang Mao and Lisa A. Karstens and Li Ma and Cindy L. Amundsen and Kenneth E. Schmader and Nazema Y. Siddiqui},
  title = {{The Urinary Microbiome in Postmenopausal Women with Recurrent Urinary Tract Infections}},
  year = {2020},
  doi = {10.1101/2020.12.21.423901},
  howpublished = {bioRxiv},
}
Figure