(A544) Evaluating the Prognostic Value of Microbial Communities in Predicting Recurrence of Laryngeal Carcinoma

Introduction: Laryngeal squamous cell carcinoma (LSCC) poses significant challenges in treatment and prognosis, particularly regarding recurrence after larynx-preservation therapy. This study investigates the relationship between microbial community composition and LSCC recurrence, focusing on the potential of specific microbiota as prognostic indicators.

Methods: We analyzed 240 formalin-fixed paraffin-embedded (FFPE) samples from 183 LSCC patients across three Chinese hospitals, including 123 patients from the LSCC cohort at the Eye & ENT Hospital, Fudan University, and 60 patients from the Multi-center cohort at two other hospitals, identifying distinct microbial profiles between recurrence and non-recurrence groups using 16S rRNA amplicon sequencing and quantitative real-time PCR (qPCR).

Results: We revealed distinct microbial community structures between recurrence and non-recurrence groups, particularly highlighting the genera abundance of Fusobacterium and Serratia. However, larynx-preserving therapy did not significantly alter microbial diversity or composition in recurrent LSCC patients. Survival analysis revealed that the amounts of Serratia and Fusobacterium, identified as independent prognostic factors for LSCC recurrence using cutoff values determined by X-tile software, led to the development of the Serratia-Fusobacterium prognostic scoring model (SF model), which achieved an area under curve (AUC) of 81.37% for predicting recurrence, outperforming the TNM staging system.LSCC patients classified as high-risk by the SF model exhibited significantly shorter disease-free survival (DFS) compared to low-risk patients in the LSCC cohort. Furthermore, the SF model achieved an AUC of 78.48% for predicting recurrence in the Multi-center cohort.

Conclusions: The Serratia-Fusobacterium prognostic scoring model can predict LSCC recurrence after larynx-preserving therapy and provide valuable insights to inform recommendations for LSCC surveillance.