Abstract / Summary
Abstract Persistent high-risk human papillomavirus (HPV) infection is the central cause of cervical cancer, but the association of the microbiome with disease severity from normal cytology to cervical lesions remains incompletely understood, particularly in African populations. Here, we characterized the cervicovaginal microbiome of 91 Kenyan women with negative cytology, low-grade, or high-grade squamous intraepithelial lesions using 16S rRNA gene sequencing, HPV genotyping, and Trichomonas vaginalis detection. Cervicovaginal communities resolved into six distinct ecological states or cervicotypes that differed in dominant taxa and composition. Increasing lesion severity was associated with depletion of Lactobacillus communities and shifts in specific bacterial taxa. Microbial co-occurrence analysis further identified Lactobacillus -, Gardnerella -, and Delftia -associated ecological modules, with the Lactobacillus -associated module inversely related to cytological severity and HPV infection. Inferred functional profiling identified 35 pathways significantly associated with increasing lesion severity, including changes in carbohydrate, amino acid, fatty acid, aromatic compound, nucleotide, and cofactor metabolism. These inferred functional signatures correlated with the Delftia -associated module. Exploratory penalized regression analyses showed that ecological features provided modest discriminatory information in addition to clinical variables. Altogether, these findings identify coordinated ecological and inferred functional changes in the cervicovaginal microbiome associated with cervical disease severity and support further investigation of microbial ecological organization in HPV-associated cervical disease.