Uterine fibroids and longitudinal profiles of the vaginal microbiota in a cohort presenting for transvaginal ultrasound

Abstract

Bacterial vaginosis, characterized in part by low levels of vaginal Lactobacillus species, has been associated with pro-inflammatory cytokines which could fuel uterine fibroid development. However, prior work on the associations between uterine fibroids and vaginal bacteria is sparse. Most studies have focused on assessment of individual taxa in a single sample. To address research gaps, we sought to compare short, longitudinal profiles of the vaginal microbiota in uterine fibroid cases versus controls with assessment for hormonal contraceptives (HCs), a possible confounder associated with both protection from fibroid development and increases in Lactobacillus-dominated vaginal microbiota. This is a secondary analysis of 83 reproductive-age cisgender women who presented for transvaginal ultrasound (TVUS) and self-collected mid-vaginal swabs daily for 1-2 weeks before TVUS (Range: 5-16 days, n = 697 samples). Sonography reports detailed uterine fibroid characteristics (N = 21 cases). Vaginal microbiota was assessed by 16S rRNA gene amplicon sequencing and longitudinal microbiota profiles were categorized by hierarchical clustering. We compared longitudinal profiles of the vaginal microbiota among fibroid cases and controls with exact logistic regression. Common indications for TVUS included pelvic mass (34%) and pelvic pain (39%). Fibroid cases tended to be older and report Black race. Cases less often reported HCs versus controls (32% vs. 58%). A larger proportion of cases had low-Lactobacillus longitudinal profiles (48%) than controls (34%). In unadjusted analysis, L. iners-dominated and low-Lactobacillus profiles had higher odds of fibroid case status compared to other Lactobacillus-dominated profiles, however these results were not statistically significant. No association between vaginal microbiota and fibroids was observed after adjusting for race, HC and menstruation. Results were consistent when number of fibroids were considered. There was not a statistically significant association between longitudinal profiles of vaginal microbiota and uterine fibroids after adjustment for common confounders; however, the study was limited by small sample size.

Fig 1. Condensed longitudinal profiles of the vaginal microbiota of 83 participants (21 cases and 62 controls).

Each column represents a participant, and each dot represents a single mid-vaginal sample classified into a CST. Condensed longitudinal profiles are presented by fibroid status (horizontal bar). The CSTs (collected daily) of each participant are shown grouped into categories of condensed longitudinal profiles. Condensed longitudinal profiles are also stratified by fibroid status.