The cervicovaginal metabolome in women with favorable induction cervix and those unfavorable for induction when delivering at term.

Background: Cervical ripening is crucial for induction. However, its influencing factors, mechanistic understanding, and effective risk stratification are still challenging. Recent research suggested that microorganisms and their metabolites in vaginal spaces correlate to preterm birth. However, it remains unclear whether the cervicovaginal metabolome is related to the natural physiological process of cervical maturation.

Objective: We aimed to analyze the cervicovaginal metabolome in women with favorable induction cervix and those unfavorable for induction when delivering at term.

Study Design: Cervicovaginal swabs were collected between 40 and 41 weeks gestation from the following 2 different groups of patients: Ripe group (n = 25) which was favorable for the induction cervix and Unripe group which was unfavorable for the induction cervix (n = 25). Samples were tested using untargeted metabolomics analysis and analyzed by a bioinformatics platform. The correlation analysis between the metabolome and the previously acquired microbiome was also performed.

Results: A total of 629 metabolites were identified in cervicovaginal fluid. The cervicovaginal metabolome was significantly different between the women with the ripe cervix and those with the unripe cervix, especially within each stratum of the same CST. Metabolites within the amino acid, carbohydrate, and dipeptide pathways may play a role in this distinction. Thirty-four metabolites were significantly upregulated, and the remaining fourteen were significantly downregulated in the Unripe group with an unripe cervix unfavorable for induction. Statistical modeling identified Arachidonic Acid and Nicotinate associated with the risk of cervical maturation disorder (AUC 0.87) in negative ion mode. A combination of Choline and d-Mannose identified a risk of cervical maturation disorder (AUC 0.80) in positive ion mode, improved by relative abundance (AUC 0.89).

Conclusion: These data suggested that the cervicovaginal space was metabolically active during pregnancy and significantly altered among the women with the mature and immature cervix. Combining the genera-level phylotypes and metabolites could build better cervix maturity prediction models. By using cervicovaginal fluid samples, we demonstrated the potential of multi-data type integration for developing composite models toward understanding the contribution of the vaginal environment to the remodeling of cervix during term pregnancy.