Oral-microbiome-derived signatures enable non-invasive diagnosis of laryngeal cancers

Abstract

Background: Recent studies have uncovered that the microbiota in patients with head and neck cancers is significantly altered and may drive cancer development. However, there is limited data to explore the unique microbiota of laryngeal squamous cell carcinoma (LSCC), and little is known regarding whether the oral microbiota can be utilized as an early diagnostic biomarker. Methods: Using 16S rRNA gene sequencing, we characterized the microbiome of oral rinse and tissue samples from 77 patients with LSCC and 76 control patients with vocal polyps, and then performed bioinformatic analyses to identify taxonomic groups associated with clinicopathologic features. Results: Multiple bacterial genera exhibited significant differences in relative abundance when stratifying by histologic and tissue type. By exploiting the distinct microbial abundance and identifying the tumor-associated microbiota taxa between patients of LSCC and vocal polyps, we developed a predictive classifier by using rinse microbiota as key features for the diagnosis of LSCC with 85.7% accuracy. Conclusion: This is the first evidence of taxonomical features based on the oral rinse microbiome that could diagnose LSCC. Our results revealed the oral rinse microbiome is an understudied source of clinical variation and represents a potential non-evasive biomarker of LSCC.