Leveraging existing 16S rRNA gene surveys to decipher microbial signatures and dysbiosis in cervical carcinogenesis.

ABSTRACT

The presence of dysbiotic cervicovaginal microbiota has been observed to be linked to the persistent development of cervical carcinogenesis mediated by the human papillomavirus (HPV). Nevertheless, the characteristics of the cervical microbiome in individuals diagnosed with cervical cancer (CC) are still not well understood. Comprehensive analysis was conducted by re-analyzing the cervical 16S rRNA sequencing datasets of a total of 507 samples from six previously published studies. We observed significant alpha and beta diversity differences in between CC, cervical intraepithelial neoplasia (CIN) and normal controls (NC), but not between HPV and NC in the combined dataset. Meta-analysis revealed that opportunistic pernicious microbes Streptococcus, Fusobacterium, Pseudomonas and Anaerococcus were enriched in CC, while Lactobacillus was depleted compared to NC. Members of Gardnerella, Sneathia, Pseudomonas, and Fannyhessea have significantly increased relative abundance compared to other bacteria in the CIN group. Five newly identified bacterial genera were found to differentiate CC from NC, with an area under the curve (AUC) of 0.8947. Moreover, co-occurrence network analysis showed that the most commonly encountered Lactobacillus was strongly negatively correlated with Prevotella. Overall, our study identified a set of potential biomarkers for CC from samples across different geographic regions. Our meta-analysis provided significant insights into the characteristics of dysbiotic cervicovaginal microbiota undergoing CC, which may lead to the development of noninvasive CC diagnostic tools and therapeutic interventions.