Discov Oncol. 2026 Feb 13. doi: 10.1007/s12672-026-04641-w. Online ahead of print.
ABSTRACT
Alterations in the vaginal microbiota contribute to the pathogenesis of cervical neoplasia. However, the distinctions in microbiota changes related to different human papillomavirus (HPV) subtypes, as well as the variation in gut microbiota, have not been fully explored. In this research, we endeavored to explore the shifts in the vaginal and intestinal microbiota in correlation with the advancement of cervical neoplasia and HPV infection. A total of 578 vaginal and intestinal cross-sectional specimens were collected from 348 subjects and subjected to 16 S rRNA sequencing. Statistical analyses were performed using R language, and Student’s t-test was employed to assess the significance of differences. Both within and between, sample diversity of the vaginal and intestinal microbiota exhibited substantial alterations across cervical intraepithelial neoplasia (CIN) stages and cervical carcinoma. The vaginal genera Lactobacillus, Enterococcus, Peptoniphilus, Atobium, Anerococcus, and Veillonella were associated with different CIN stages and cervical cancer type, whereas Allisonella, Lachnospiracae, Lactobacillus, Staphylococcus, and Sellimonas were associated with varying HPV types. A Random Forest-driven classifier highlighted the predictive potential of differential bacteria in cervical neoplasia and HPV infection, with intestinal bacteria showing higher predictive accuracy in certain instances. Specifically, the accuracy of differentiating CIN I from CIN III was superior for the intestinal bacterial model compared to the vaginal bacterial model (85.52% vs. 83.33%). The model also demonstrated high accuracy in predicting HPV infection, particularly in distinguishing HPV-16 from HPV-18 and HPV-58, with AUC values of 81.61% and 83.07%, respectively, compared to less than 70% for vaginal bacteria. Our findings reveal the intricate interplay among cervical neoplasia, HPV infection, and microbiota, with potential diagnostic and therapeutic implications.
PMID:41686340 | DOI:10.1007/s12672-026-04641-w
