The Microbial Composition of Lower Genital Tract May Affect the Outcome of in vitro Fertilization-Embryo Transfer

Objective: This work was conducted in order to study the effect of the lower genital tract (vaginal and cervical canal) microbiota on pregnancy outcomes of reproductive-aged women receiving embryo transfer. Study design: A total of 150 reproductive-aged patients who received the first fresh in vitro fertilization-embryo transfer (IVF-ET) were included in the study. Samples from the vagina and cervical site of each patient were collected separately using sterile swabs before ET. Genomic DNA was pyrosequenced for the V3-V4 regions of the 16S ribosomal RNA gene. Further bioinformatics analysis was performed using QIIME and R package. Pregnancy outcomes were followed and analyzed to compare differences in microbial composition. Results: The cervical microbiota had a higher Shannon index than the vaginal microbiota, and the microbial composition was different between the two sites. However, the Sorenson index between the two sites within the same individual was 0.370 (0.309-0.400). A total of 89 patients achieved clinical pregnancy after ET, while 61 failed. The Shannon indices and the microbial community of both vaginal and cervical microbiota between pregnant and non-pregnant groups were not significantly different. The relative abundance of Lactobacillus in the vagina and cervical canal did not differ between the two groups. Linear discriminant analysis, random forest analysis, and receiver-operating characteristic curve analysis showed that Bifidobacterium, Prevotella, and Lactobacillus iners in the vagina, as well as Solanum torvum, Fusobacterium, and Streptococcus in the cervix, may be negatively associated with clinical pregnancy after IVF. Conclusion: The cervical microbiota was more diverse than the vaginal microbiota, but because of anatomical continuity, there was a correlation between the two sites. The microbial composition of the vagina and cervical canal may influence the outcome of IVF-ET, but more samples are needed to verify this conclusion.