Abstract
The aim of this study was to evaluate changes in the vaginal microbiota and biomarkers among high-risk human papillomavirus (hrHPV)-positive women, those with hrHPV accompanied by mucositis, and patients with cervical intraepithelial neoplasia (CIN) and to establish a novel predictive model. Vaginal samples from 102 women were categorized into four groups: control group (n = 26), hrHPV-positive group (n = 22), hrHPV-positive with mucositis group (n = 26), and CIN group (n = 28). Microbiota analysis was performed using the PacBio platform with full-length 16S rDNA gene sequencing. The vaginal microbiota in the hrHPV-positive, hrHPV-positive with mucositis, and CIN groups showed significant differences compared with the healthy control group. The microbial richness in the hrHPV-positive group was significantly different from both the CIN group and healthy controls. Compared with the control group, the hrHPV-positive group exhibited significantly increased relative abundances of Bifidobacterium, Escherichia-Shigella, Hoylesella and nominally increased abundances of Gardnerella, Prevotella, along with a significant decrease in Lactobacillus. No statistically significant differences were retained between the hrHPV-positive group and the hrHPV-positive with mucositis group after FDR correction for the top 10 genera. Compared with the hrHPV-positive with mucositis group, the CIN group demonstrated significantly reduced levels of Pseudomonas, nominally decreased levels of Bifidobacterium and Faecalibacterium, whereas Glutamicibacter and Sporosarcina were nominally enriched. A random forest model was constructed to predict risk across groups and demonstrated good predictive performance, suggesting that vaginal microbiota may serve as valuable indicators for predicting cervical lesion risk. During hrHPV infection, significant alterations occur in the vaginal microecology, primarily characterized by an increase in pathogenic bacteria and a reduction in beneficial bacterial populations.