Describe pre-operative urinary and vaginal bacterial taxa associated with 12 month surgical response after mid-urethral sling in women with mixed urinary incontinence.

After Institutional Board Approval and written informed consent was obtained, urine and vaginal bacterial samples (N=124) were collected from a subset of women with MUI enrolled in the ESTEEM trial (NCT01959347). All women underwent midurethral sling surgery. Objective response at 12 months was defined as ≥70% reduction from baseline in urinary incontinence episodes on 3-day diary without additional treatment. Based on mixed urinary incontinence symptoms as measured by the Urogenital Distress Inventory, as most subjects responded to treatment meeting the minimal clinically important difference threshold, subjective response was defined by change from baseline in Urogenital Distress Inventory analyzed continuously. Clinical and demographic differences in responder status were assessed with univariate generalized linear models and Fischer’s exact test. Using 16S rRNA gene sequencing data, beta diversity was compared in responders vs non-responders. Differential abundance analysis of operational taxonomic units was assessed, focusing on predominant taxa defined as the most prevalent operational taxonomic units in ≥3 participants, using unadjusted and age-adjusted linear models.

Objective responders (N=72), compared to non-responders (N=28) were younger (51.6±10.2 y, p = 0.007), premenopausal (OR [CI] 4.22 [1.28-13.86]) and on hormone replacement therapy (HRT), (OR [CI] 6.35 [1.82-22.23]. Age (urine and vagina, p<0.001), menstrual status (urine, p=0.04; vagina, p=0.01), postmenopausal +HRT (vagina, p=0.01) and responder status (vagina, p=0.03) were associated with operational taxonomic units in beta diversity analyses. Six predominant operational taxonomic units were identified: Lactobacillus, most predominant (urine and vagina), followed by Gardnerella (urine and vagina), Tepidomonas (urine), Escherichia (urine), Streptococcus (urine and vagina), Prevotella (urine and vagina). Differential abundance analysis of urine revealed no association between predominant operational taxonomic units and responder status. Vaginal differential abundance analysis adjusting for age, found Prevotella associated with objective responder status, p=0.01; changes in Urogenital Distress Inventory (N=105) scores were not associated with vaginal operational taxonomic units (Fig. 1a, b). Lower Lactobacillus was associated with older age and higher Prevotella (Fig. 2).

In older women undergoing a mid-urethral sling for the treatment of mixed urinary incontinence, the state of the vaginal microbiome may be important in optimizing surgical outcomes.

Surgical responders had decreased Prevotella in the vagina. Younger women, more likely to be surgical responders, had more Lactobacillus. Further research is needed to confirm whether therapy altering the vaginal microbiome in older women impacts surgical response in women with mixed urinary incontinence.