Microbiome characteristics associated with lymph node metastasis in laryngeal squamous cell carcinoma.

Lymph node (LN) metastasis is a key prognostic factor in laryngeal squamous cell carcinoma (LSCC). Emerging evidence implicates the role of the microbiome in cancer progression. This study aimed to investigate the microbial features associated with lymph node metastasis in LSCC and their potential for patient stratification. Using 16 S rRNA gene sequencing, we characterized the microbiome of tumor tissues, adjacent normal tissues, lymph nodes, and oral rinses from 108 LSCC patients, including 36 with (LN+) and 72 without (LN-) cervical LN metastasis. Microbial functional potential was predicted using PICRUSt2. Based on repeated stratified 3 cross-validation, random forest models were used to identify metastasis-associated genera. Significant microbial differences were observed between LN + and LN- tumor tissues, with Ralstonia enriched in LN + tumors and Fusobacterium more abundant in LN- cases. All genera detected in lymph nodes were also found in tumor tissues. Functional predictions revealed enrichment of lipid biosynthesis, energy metabolism, and cell wall synthesis pathways in LN + patients, particularly in tumor and oral rinse samples, with low intra-group variability. Classifiers based on tumor, lymph node, and oral microbiota demonstrated the ability to distinguish LN + from LN- patients. The lymph node-derived classifier achieved an accuracy of 84.31% (95% confidence interval [CI]: 81.76% - 86.85%), followed by the tumor-based model (AUC = 84.11%, 95% CI: 81.75% - 86.46%) and oral rinse classifier (AUC = 79.88%, 95% CI: 77.09% - 83.11%). A tumor-specific 17 genera panel showed a discriminative efficacy of 84.11% (95% CI: 81.75% - 86.46%) in tumor tissues. These findings suggest that microbiome alterations may be associated with lymph node metastasis in LSCC. In addition, the oral microbiome showed potential as a non-invasive tool for occult lymph node metastasis detection. However, these results are preliminary and require validation in larger, independent cohorts.