# A potential association between the characteristics of the multi-organ microbiota and lymph node metastasis in cervical cancer

## TL;DR

This study explores how changes in microbiota across multiple organs may be linked to lymph node metastasis in cervical cancer patients.

## Contribution

The paper introduces a noninvasive prediction model based on oral microbiota to assess lymph node metastasis in cervical cancer.

## Key findings

- The urinary microbiota showed reduced α-diversity in patients with lymph node metastasis.
- A predictive model using oral bacteria achieved 94.1% specificity and 57.9% sensitivity for metastasis detection.
- Significant differences in gut and urine microbiota structure were observed between metastatic and non-metastatic groups.

## Abstract

Microbiota alterations at multiple sites are associated with cervical cancer (CC). However, it is unclear whether CC lymph node metastasis (LNM) is indeed associated with microbiota alterations, whether the microbiota is generally suitable for screening CC LNM-related taxa.

We performed 16S rDNA sequencing of samples from oral swabs, feces, urine, and vaginal secretions from CC patients to clarify microbiota characteristics of LNM group. And we constructed a LNM prediction model for CC based on specific flora at each site.

The α-diversity of the urinary microbiota (PSob = 0.0272, PPielou = 0.0278, PShannon = 0.0209 and PSimpson = 0.0465) was reduced in the LNM group compared to the non-LNM group, and significant differences were observed in the structure of the gut (R² = 0.0266, P = 0.033) and urine (R² = 0.0379, P = 0.002) microbiota between the two groups. The establishment of a predictive model based on oral specific flora, including Erysipelotrichaceae UCG-003 sp., Eubacterium halli group, and Staphylococcus has enabled the differentiation of CC lymph node status. The area under the ROC curve was 0.798. The Yoden index, sensitivity and specificity of this prediction model were 0.520, 57.9% and 94.1%, respectively.

CC patients with LNM have significant microbiological changes at multiple sites. The predictive model based on oral bacteria can provide a noninvasive and simple method for assessing LNM in CC.