Role of the vaginal microbiome in preterm prelabour rupture of the membranes: an observational study. (25th February 2016)

• Record Type: Journal Article
• Authors: Brown, Richard; Kindinger, Lindsay; Lee, Yun S; Marchesi, Julian R; Bennett, Phillip; MacIntyre, David
• Is Part Of: Lancet. Volume 387(2016)Supplement 1
• Page Start: S22
• Publication Date: 2016-02-25
• Subjects: Medicine -- Periodicals; Medicine -- Periodicals; Medicine; Medicine; Electronic journals; Periodicals; 610.5
• DOI: 10.1016/S0140-6736(16)00409-8
• Languages: English
• ISSNs: 0140-6736

Abstract

Abstract: Background: Around 30% of preterm births are preceded by prelabour rupture of the membranes, and pathobionts ascending from the vagina are implicated in 40% of cases. A detailed understanding of the vaginal microbiome in preterm prelabour rupture of the membranes (PPROM) is pivotal for the development of predictive, preventive, and therapeutic strategies. We aimed to assess the vaginal microbiome of women with PPROM and the effect of erythromycin treatment. Methods: Vaginal swabs were taken from pregnant women at four antenatal timepoints and following PPROM before and after erythromycin treatment. A subset of those women sampled antenatally subsequently developed PPROM whereas the remainder delivered at term, forming the control group from which gestational age-matched (27–30 weeks') samples were drawn. Swabs were immediately frozen and stored at −80°C, and bacterial DNA was extracted by enzymatic and mechanical disruption. Bacterial species composition and structure of resident communities were examined by next generation sequencing of 16S rRNA gene amplicons, V1-V3 (Illumina MiSeq, San Diego, CA, USA). Findings: 91 pregnant women were examined. Women who went on to develop PPROM (n=12) had a vaginal microbiome before membrane rupture that was dominated by Lactobacillus iners (7, 58%), L jenseni (1, 8%), and L crispatus (1, 8%); three women (25%) had a vaginal microbiome low in Lactobacillus spp and enriched with Gram-negative bacteria such as Prevotella spp (dysbiotic). Control women (n=39) had a vaginal microbiome dominated by L crispatus (17, 43%), L iners (12, 31%), L jenseni (5, 13%), and L gasseri (5, 13%), with no samples displaying dysbiosis. After membrane rupture but before erythromycin treatment (n=17), the vaginal microbiome was dominated by L iners (8, 47%), dysbiosis (5, 29%), L crispatus (3, 18%), and L iners plus L crispatus (1, 6%). After 48 h of erythromycin treatment (n=23), dysbiosis increased to become the dominant state in 18 women (74%). Interpretation: Our finding that a vaginal microbiome devoid of L crispatus before membrane rupture is a feature of pregnancies complicated by PPROM supports a protective role for L crispatus in the prevention of this complication of pregnancy. Dysbiosis rises after PPROM and becomes the dominant community state after erythromycin treatment. This finding suggests that environmental perturbations of membrane rupture, in combination with erythromycin treatment, promotes dysbiosis and a community state that harbours bacteria implicated in maternal and neonatal sepsis. Funding: Medical Research Council.