Multi-omic analysis of vaginal bacteria:host immune and metabolic interactions in adverse early pregnancy outcomes
Accurate and timely diagnosis of early pregnancy outcomes remains a clinical challenge, particularly in cases classified as pregnancy of unknown location, where ultrasound fails to confirm the location of the pregnancy. Delayed diagnosis can lead to serious complications, especially in ectopic pregnancies. This prospective cohort study aimed to characterise host-microbe interactions during early pregnancy in the vaginal mucosa by exploring the relationship between the bacterial community composition, immune response, and metabolic phenotype. We analyzed vaginal swabs from 91 women with pregnancy of unknown location, including 22 with viable intrauterine pregnancies and 69 with adverse outcomes. Metataxonomic profiling, immune assays, and direct on-swab metabolomics were performed. Adverse outcomes, especially ectopic pregnancies, were associated with reduced Lactobacillus abundance, increased bacterial diversity, mucosal inflammation, and elevated lipid species. Vaginal bleeding, a common symptom in PUL, significantly impacted immune and metabolic profiles and should be considered as a confounder in studies of the lower female reproductive tract. Multi-omic data integration highlighted metabolic phenotypes of immune response in the context of bacterial community composition, underscoring the influence of microbe-host interactions on mucosal health and pregnancy outcomes. IMPORTANCEEarly detection of complications in pregnancy is essential for improving maternal health outcomes, yet current diagnostic tools often fail to identify ectopic or failing pregnancies in a timely manner. This study highlights how changes in the vaginal microbiome, inflammation, and metabolic signals are closely linked to adverse outcomes in early pregnancy. By showing that metabolic profiling can predict microbial and immune states even in the presence of vaginal bleeding, the research provides a foundation for developing rapid, non-invasive diagnostic tools. These insights could pave the way for more personalized and effective early pregnancy care.