A nomogram prediction model for embryo implantation outcomes based on the cervical microbiota of the infertile patients during IVF-FET.

The microbiota of the female genital tract is crucial for reproductive health. This study aims to investigate the impact of the lower genital tract microbiota on fertilization and frozen embryo transfer (IVF-FET) outcomes. This study included 131 women aged 20-35 years who underwent their first or second IVF-FET cycle with no obvious unfavorable factors for implantation. Cervical microbiota samples were collected on the embryo transfer day and analyzed using 16S rDNA full-length sequencing. Clinical outcomes were followed up for analysis. Clinical pregnancy (CP) was achieved in 84 patients, and 47 patients experienced non-pregnancy (NP). The cervical microbiota diversity between NP and CP groups showed no significant differences, but some genera such as ( = 0.018), ( = 0.039), ( = 0.016), and ( = 0.021) were obviously different between the two groups. Notably, there was no significant difference in the abundance of between the two groups. A nomogram prediction model was developed using the random forest algorithm and logistic regression, including the classification of , , and , as well as the relative abundance of , to identify high-risk patients with embryo implantation failure. Both internal (area under the curve [AUC] = 0.718, 95% confidence interval [CI]: 0.628-0.807, < 0.001) and external validation (AUC = 0.654, 95% CI: 0.553-0.755, = 0.037) of the model performed well. In conclusion, this study established a correlation between cervical microbiota and embryo implantation failure in infertile women undergoing IVF-FET and developed a prediction model that could help in early identification of patients at high risk of implantation failure.IMPORTANCEThis study investigated the potential role of abnormal cervical microbiota in the pathology of implantation failure after fertilization and frozen embryo transfer (IVF-FET) treatment. Despite nearly half a century of advancements in assisted reproductive technology (ART), the implantation rate of high-quality embryos still hovered around 50%. Moreover, unexplained recurrent implantation failure (RIF) remains a significant challenge in ART. To our knowledge, we first discovered a prediction model for embryo implantation failure, identifying and as significantly adverse factors for embryo implantation. Despite some limitations, the internal and external validation of the model could bode well for its clinical application prospect. The insights gained from this study pave the way for intervention in the genital tract microbiota prior to IVF-FET, particularly in patients with RIF and RSA.