15: Gardnerella vaginalis and spontaneous preterm birth: New insights


      Despite demonstrated epidemiological associations between bacterial vaginosis (BV) and spontaneous preterm birth (sPTB), clinical trials targeting eradication of BV have failed to reduce sPTB. We hypothesize that the diagnosis of BV (by clinical criteria or Nugent) is non-specific and does not address specific bacteria, interactions among bacteria or the possible immune responses to these bacteria in determining risk for sPTB. We investigated whether relative abundance (RA) of Gardnerella vaginalis (G. vag) was associated with sPTB and if immune responses in the cervicovaginal (CV) space modify any association.

      Study Design

      CV biospecimens were collected at three time points in a prospective cohort of singleton pregnancies (“M&M”; n=1,943). A nested case-control study of 405 term and 98 sPTB was performed using biospecimens collected at 16-20 weeks. Two antimicrobial peptides (AMPs) (Beta Defensin (BD) and SLPI) were assessed by ELISA and microbiota RA was determined with 16s rRNA analysis. Maternal characteristics and pregnancy outcomes were examined by RA quartile of G. vag. Odds of sPTB by G. vag quartile was estimated with logistic regression adjusting for race, pH >5 at time of visit, intercourse ≤24 hrs, BD, and quartile of Lactobacillus crispatus (L. crisp) and Mobiluncus mulieris (MM). Sensitivity analyses examining race and immune markers were also performed.


      G. vag was present in 98.6% of samples. In univariate analysis, there were significant differences across quartiles in race, basic pH, BD, SLPI, and RA of L. crisp and MM, but not in the rate of sPTB (p=0.07). After adjustment, G. vag was associated with elevated risk of sPTB at Q3: aOR: 3.1 (1.4-6.8). When stratified by race, the same association was seen at Q3, but was stronger among white women than African American (AA) women (aOR: 16.3 vs. aOR: 3.0). Level of BD appears to mediate the impact of G. vag. (fig)


      Clinical trials targeting BV likely failed because they did not address other microbiological and immunological factors in the CV space that modify the risk of sPTB in the presence of bacteria known to be associated with BV. Higher RA of G. Vag and low BD level confers a high rate of sPTB. Higher RA of G. vag is much more common in AA women but non-AA women at are higher risk. Increased understanding of the microbe-immune interactions in the CV space by race will likely yield more targeted therapeutics to reduce sPTB.
      Figure thumbnail fx1
      Figure thumbnail fx2