Potential Association between Vaginal Microbiota and Cervical Carcinogenesis in Korean Women: A Cohort Study

Abstract

Convincing studies demonstrated that vaginal flora is one of the most impactful key components for the well-being of the genital tract in women. Nevertheless, the potential capability of vaginal-derived bacterial communities as biomarkers to monitor cervical carcinogenesis (CC) has yet to be studied actively compared to those of bacterial vaginosis (BV). We hypothesized that vaginal microbiota might be associated with the progression of CC. In this study, we enrolled 23 participants, including healthy controls (HC group; n = 7), patients with cervical intraepithelial neoplasia (CIN) 2 and 3 (CIN group, n = 8), and patients with invasive cervical cancer (CAN group; n = 8). Amplicon sequencing was performed using the Ion Torrent PGM to characterize the vaginal microbiota. Patients with CIN and CAN presented vaginal microbiota dysbiosis compared with HC. The alpha diversity analysis revealed that CC has a trend to be increased in terms of diversity indexes. Moreover, CC was associated with the abundance of specific microbes, of which Lactobacillus and Gardnerella were the most significantly different between HC and CIN, whereas Streptococcus was differentially abundant in CAN compared with CIN. We then evaluated their diagnostic abilities. Testing in terms of diagnostic ability using the three genera revealed considerably high performance with an area under the receiver-operating characteristic curve of 0.982, 0.953, and 0.922. The current study suggests that the presence of Gardnerella and Streptococcus may be involved in the advancment of CC.

Keywords: CIN prediction; CIN severity; vaginal microbiome; vaginosis.

Figure 1

The differences in vaginal microbiota diversity by health status. (A) The boxplots display the differences in the alpha diversity indexes (Shannon index: left panel, Richness: middle panel, and Simpson’s index: right panel) between the groups. (B) Principal coordinate analysis (PCoA) plot depending on the Bray–Curtis dissimilarity of beta diversity, which colored each sample according to health status. The HC group shows lower inter-individual variations than the CIN and CAN groups at the PCoA axis1; * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 2

Comparative analysis of vaginal microbiota profiles. Differences in microbial composition between each group at the genus (A,C) and corresponding phylum levels (B,D) are presented as a heat map. The LDA effect size (LEfSe) analysis to identify the potential biomarkers revealed changes of the vaginal microbiota according to health status (E,F). (G) The Bray–Curtis dissimilarity based PCoA plot is colored by the relative abundance of the Lactobacillus and Streptococcus genera; the left hemisphere indicates the health status and the right presents the relative abundance of each bacterial genus. Statistical significance was computed using the Wilcoxon rank–sum test; * p < 0.05, ** p < 0.01, *** p < 0.001.

Figure 3

Evaluation of the top five bacterial genera as a noninvasive diagnostic tool to predict disease progression. (A) ROC analysis to compare HC with CIN. (B) ROC analysis to compare CIN with CAN. The left panel displays the AUC of the top five impactful genera, and the right panel presents the individual AUC of the ROC curves.

Figure 4

Vaginal microbiota composition and their correlation pattern analysis comparing HPV positive to negative. (A) Heat map of the differentially abundant microbiota profile between HPV positive and negative (Wilcoxon sum test; only the genera with p < 0.05 were selected). (B) Alpha diversity including the Shannon index, Richness, and Simpson’s index is plotted for participants with HPV negative (pink) and positive (green). (C) Spearman correlation coefficients (Rho) are presented red for positive correlations and blue for negative correlations. The intensity of the color and size of the dot are associated with the Rho coefficient strength. (D) Scatter plots with illustration of the correlation between the specific bacterial genus and alpha diversity of each group’s vaginal microbiota.