Introduction. Multiple reports have attempted to describe the tumour microbiota in head and neck cancer (HNSC).
Gap statement. However, these have failed to produce a consistent microbiota signature, which may undermine understanding the importance of bacterial-mediated effects in HNSC.
Aim. The aim of this study is to consolidate these datasets and identify a consensus microbiota signature in HNSC.
Methodology. We analysed 12 published HNSC 16S rRNA microbial datasets collected from cancer, cancer-adjacent and non-cancer tissues to generate a consensus microbiota signature. These signatures were then validated using The Cancer Microbiome Atlas (TCMA) database and correlated with the tumour microenvironment phenotypes and patient’s clinical outcome.
Results. We identified a consensus microbial signature at the genus level to differentiate between HNSC sample types, with cancer and cancer-adjacent tissues sharing more similarity than non-cancer tissues. Univariate analysis on 16S rRNA datasets identified significant differences in the abundance of 34 bacterial genera among the tissue types. Paired cancer and cancer-adjacent tissue analyses in 16S rRNA and TCMA datasets identified increased abundance in
Fusobacterium
in cancer tissues and decreased abundance of
Atopobium
,
Rothia
and
Actinomyces
in cancer-adjacent tissues. Furthermore, these bacteria were associated with different tumour microenvironment phenotypes. Notably, high
Fusobacterium
signature was associated with high neutrophil (r=0.37, P<0.0001), angiogenesis (r=0.38, P<0.0001) and granulocyte signatures (r=0.38, P<0.0001) and better overall patient survival [continuous: HR 0.8482, 95 % confidence interval (CI) 0.7758–0.9273, P=0.0003].
Conclusion. Our meta-analysis demonstrates a consensus microbiota signature for HNSC, highlighting its potential importance in this disease.