852: Longitudinal metagenomic survey of vaginal Group B Strep (GBS) status and microbial community structure suggests transient culture sensitivity

Abstract

The current clinical standard for detection of GBS (Streptococcus agalactiae) colonization is vaginal rectal cultures; however, since PCR-based tests are used screening, it is possible that metagenomic approaches might provide a more sensitive assessment of GBS status. Given that metagenomics is the gold standard for DNA typing of bacterial species/strains, our study aim was to determine the concordance between vaginal metagenomics and GBS culture. In this prospective longitudinal study, we performed whole genome shotgun (WGS) metagenomic sequencing on vaginal samples (n=164) collected from gravidae (n=60 subjects) during the 3rd trimester, at delivery & at 4-6 weeks postpartum; all subjects underwent culture-based GBS screening at 36 weeks. Two vaginal sites were sampled (introitus & posterior fornix), and extracted DNA was subjected to WGS sequencing to speciate & quantify GBS abundance. 108 samples (∼1.9 billion reads) were GBS positive by WGS sequencing (WGS-GBS-pos), and 17% of all subjects were negative by both culture and WGS sequencing. All culture positive subjects (n=5) were WGS-GBS-pos at all sampling times, compared to 82% (n=45/55) of culture negative subjects. Markov modeling (1A) demonstrates that both WGS-GBS-pos and WGS-GBS-neg status have a high probability of becoming or staying WGS-GBS-pos. WGS-detected GBS did not result from DNA contamination, given high relative abundance in culture-negative samples (1B). Vaginal communities were significantly different by virtue of culture positivity (1C,D), and antibiotic treatment at delivery did not reduce the relative abundance of GBS postpartum. ROC analysis as well as calculated sensitivity, specificity, NPV and PPV were consistent with improved prediction by sampling of the vaginal introitus relative to the posterior fornix and by presence or absence of GBS metagenomes at multiple sampling points (2). These data demonstrate 100% concordance of vaginal metagenomics with positive GBS culture. Our findings suggest that GBS prevalence by multiple positive WGS samplings would detect an additional 20-40% of culture negative subjects for an overall prevalence nearing 50%. Collectively, these data suggest that while evidence of GBS colonization by culture may be temporal & transient, detection by WGS is not.View Large Image Figure ViewerDownload Hi-res image Download (PPT)