The cervicovaginal microbiota has co-evolved with mucosal immune cells to facilitate fetal implantation and development during pregnancy and protect the female urogenital tract from infection (1, 2). The correlation between cervicovaginal microbiota and its role in either fostering susceptibility or providing protection against HPV, the third leading cause of cervical cancer, is gaining attention (3).

Research on bacterial communities has revealed that the optimal vaginal microbiota is predominantly composed of Lactobacillus crispatus, L. gasseri, and L. jenseni. These species secrete lactobiocins and biosurfactants that prevent other species from binding to the epithelium (4). Interestingly, a vaginal microbiota dominated by L. iners has been associated with higher odds of HPV prevalence, including high-risk HPV and cervical cancer or dysplasia (5). It has been established that the interaction between HPV, host immunity, and the vaginal microenvironment influences the risk of persistent HPV infection and the progression of cervical intraepithelial neoplasia (CIN) (6, 7). However, while HPV-related complications primarily develop in the cervix, the cervical microbiota has been less studied. Comparative analyses not focused on HSIL further show cervix–vagina differences, underscoring site-specific composition and detectability, and reporting variable Lactobacillus dominance (8, 9). Furthermore, decreased levels of Lactobacillus have been reported in women with low HPV viral loads (10).

Therefore, we conducted a study to explore the connections between HPV infection and squamous intraepithelial lesions, focusing on the composition and function of the cervical microbiota. We aimed to identify metabolites that can help identify risk markers for malignant or premalignant lesions, enhance current screening efforts, and identify potential therapeutic targets in the long term.

Our study revealed a more diverse vaginal microbiota in women with cervical dysplasia than that in healthy women, with differences in specific taxa and biological functions linked to the development of cervical HSIL. Increased microbial diversity in the cervical environment, as observed in women with HSIL, may reflect a shift away from a Lactobacillus-dominated state toward a polymicrobial community enriched in anaerobes and associated with local inflammation. Such transitions have been linked to epithelial barrier disruption, altered immune responses, and HPV persistence, all of which may contribute to lesion progression (12, 13). The increased abundance of Parvimonas, Fastidiosipila, and Pseudomonas in women with HSIL suggests their potential role in cervical dysbiosis and HPV progression. Notably, Parvimonas spp. has been frequently associated with precancerous lesions and dysbiosis (7), particularly in connection with sexually transmitted infections such as Chlamydia trachomatis and persistent HPV infection. Interestingly, Pseudomonas, while often considered part of the healthy endocervical microbiota, has been linked to various pathologies such as Chlamydia trachomatis infection in some in vitro studies (14, 15), further supporting the idea that shifts in the microbiota may contribute to disease. The prominence of these genera, which are more commonly associated with the intestinal microbiota, suggests possible translocation or shared pathways between the gut and the cervicovaginal environment (16, 17). Mechanisms such as bacterial ascent through the reproductive tract or hematogenous spread from gastrointestinal niches could explain this crossover. However, further studies are required to explore these possibilities.

The functional analysis further reinforces the association between microbial dysbiosis and HSIL. We found enriched gene pathways related to typical bacterial metabolic processes and enzyme pathways related to glycine, serine, and threonine metabolism, suggesting that the microbial community in HSIL cases exhibits heightened metabolic activity. The increase in glycosidases and a corresponding decrease in oxidoreductases also point to significant metabolic alterations, potentially contributing to a dysbiotic environment conducive to lesion progression. These findings are consistent with metabolic dysregulation linked to cancer development, including processes such as the Warburg effect and oxidative stress, which may promote DNA damage and promote the development of cancerous lesions (18, 19).

Our study has some limitations. First, although we observed robust and consistent associations across diversity indices, taxonomic composition, and functional pathways, the relatively small number of HSIL cases (n = 18) and the amplicon-based sequencing technique may have limited our ability to detect more subtle differences. Nonetheless, the effect sizes observed were sufficient to reach statistical significance despite this limitation. Second, certain behavioral and hormonal factors known to influence the cervicovaginal microbiota—such as contraceptive use, recent antibiotic exposure, or sexual behavior—were not systematically collected, which may represent sources of residual confounding. While our findings are representative of a well-defined Spanish population, we acknowledge the need for larger, multicenter, and multiethnic studies to confirm and extend these results. Importantly, the microbial shifts we observed are consistent with reports from other settings, supporting the biological plausibility and relevance of our findings. Although our study was not designed to assess the clinical utility of microbiota-based markers, the microbial and functional differences observed may help inform future efforts to develop adjunctive tools for risk stratification. We have previously shown the predictive value of microbiome and metabolomic signatures for anal HSIL progression in a prospective cohort (20), and similar strategies are now being planned for cervical dysplasia.

Microbiome-based interventions, particularly intravaginal administration of Lactobacillus crispatus and long-term use of Lactobacillus rhamnosus, have demonstrated potential in improving cervical cytology and promoting high-risk HPV clearance in small randomized or controlled trials (20, 21). While these findings are promising, they remain limited by small sample sizes and short follow-up durations. Larger, rigorously designed randomized controlled trials are urgently needed to confirm efficacy and guide clinical implementation. Additionally, most studies examining the relationship between microbiota and cervical dysplasia have focused on vaginal rather than cervical samples (21) or have reported mixed results from different anatomical sites (22). Future studies correlating both environments could provide a less invasive means of assessing women at a higher risk of progressing to HSIL by linking findings in the vaginal microbiota to those in the cervix.

Geographic location may also influence microbiota composition (23). To the best of our knowledge, this study is the first to examine the link between the cervical microbiota and HSIL in women from Spain, and direct comparisons with the existing literature are challenging. This underscores the need for region-specific studies to better understand the microbial landscape and its relationship to cervical dysplasia. Lastly, although our functional analysis identified significant pathways, the presence of genes does not guarantee their activity, highlighting the importance of further experimental studies, particularly longitudinal studies, to establish causal relationships between cervical dysbiosis and CIN progression.

We acknowledge that cross-sectional studies such as ours cannot determine causality. Although key exclusion criteria were applied to reduce major confounders, we acknowledge that many other factors influencing microbiota composition—such as diet, probiotic use, or menstrual cycle phase—were not systematically collected and represent a limitation of this study. In our regrouped analysis (HSIL vs. non-HSIL), being currently in a couple was more frequent among women with HSIL, however, while this finding may reflect unmeasured aspects of sexual behavior or partner-related factors, it should be interpreted cautiously given the lack of detailed behavioral data. These limitations have been broadly discussed in our recent review (24), where we emphasized the need for complementary experimental models —such as experimental models, multi-omics approaches, and longitudinal cohort designs—to better elucidate the causal relationships between the microbiota and viral pathogenesis. Although we observed differences in variables such as smoking, age, or nationality across groups, our study was not designed or powered to disentangle their individual effects on microbiota composition. Future hypothesis-driven studies with larger sample sizes and multivariable modeling will be required to evaluate the independent contribution of these host factors.

Despite these limitations, our study lays the groundwork for future research to explore the role of the cervicovaginal microbiome in HSIL development. The microbial taxa and metabolic pathways identified here could inform studies focused on identifying markers for cervical dysplasia progression, offering avenues for targeted screening and therapeutic interventions.