Abstract
BACKGROUND: Cervical cancer (CC) is the fourth most prevalent malignancy among women worldwide, where 99.7% of the cases are linked to persistent human papillomavirus (HPV) infections. While emerging evidence suggests a role for vaginal microbiome dysbiosis in HPV-driven CC, the specific microbial alterations and their functional implications remain unclear. However, inconsistencies in identifying specific microbial signatures-largely due to heterogeneous study designs, targeted 16S rRNA regions, and data processing methods-have limited the generalizability of existing findings. To address these challenges, we conducted a standardized mega-analysis using a compositionality-aware approach to ensure consistency and minimize technical bias across studies. RESULTS: Our mega-analysis consolidates findings from five case-control 16S rRNA ampilicon sequencing studies, encompassing 215 samples. Compared to healthy controls, CC patients exhibited significantly higher alpha diversity (Shannon index, p <0.005) and a shift from a Lactobacillus-dominant to a polymicrobial vaginal microbiome. This microbial dysbiosis was characterized by an increased abundance of Porphyromonadaceae, particularly Porphyromonas asaccharolytica, and other anaerobic bacterial species such as Campylobacter ureolyticus, Peptococcus niger, and Anaerococcus obesiensis (FDR <0.05). Functional profiling of the altered microbiome revealed enrichment in pathways associated with chronic inflammation, fatty acid biosynthesis, amino acid metabolism, cellular proliferation, invasion, and metastasis. CONCLUSIONS: This mega-analysis presents the most methodologically homogeneous study to date of CC-associated vaginal microbiome using publicly available 16S datasets. Our findings not only deepen our understanding of microbial influences on CC but also pave the way for novel diagnostic and therapeutic approaches potentially enhancing patient outcomes in CC care. These insights open new avenues for clinical interventions that extend beyond conventional HPV-centric strategies.