ESTRO 2025 - Abstract Book

S1126

Clinical – Head & neck

ESTRO 2025

In training, the classification tree model (Figure 2A) reached an overall accuracy of 76% (sensitivity 92%, specificity 68%, McNemar's test p <001). When the mixed subgroup of patients (n=73) was stratified using the model, the overall survival of patients predicted to be hypoxia-high/immune-low showed a trend for unfavourable outcome compared to hypoxia-low/immune-high predicted patients (median survival 2.7 years versus 4.7 years, Figure 2B).

Conclusion: We demonstrated that information on the bacterial composition of HNSCC tumour microenvironment could be exploited and integrated with tumour genomics/transcriptomics data to improve prognosis.

Keywords: microbiota, hypoxia, machine learning

References: [1] Sambasivan et al. TNM 8 staging is a better prognosticator than TNM 7 for patients with locally advanced oral cavity squamous cell carcinoma treated with surgery and post-operative radiotherapy. Radiotherapy and Oncology (2021). [2] Perez-Chanona E., Trinchieri G. The role of microbiota in cancer therapy. Current opinion in immunology 39 (2016). [3] Brooks J. M. et al. Development and validation of a combined hypoxia and immune prognostic classifier for head and neck cancer. Clinical Cancer Research (2019) [4] TCGA Research Network: https://www.cancer.gov/tcga [5] Dohlman A.B. et al. The cancer microbiome atlas: a pan-cancer comparative analysis to distinguish tissue resident microbiota from contaminants. Cell host & microbe (2021).

4261

Digital Poster incidence of nodal involvement of IX level in Head & Neck cancer: a multicenter retrospective analysis. Fabrizio Sanna 1 , Anna Rita Musu 2 , Anna Merlotti 3 , Dalila Di Mario 1 , Claudia Crescio 4 , Francesco Bussu 4 , Salvatrice Campoccia 1 1 Radiotherapy, AOU Sassari, Sassari, Italy. 2 Radiotherapy, Asl Nuoro, Nuoro, Italy. 3 Radioterapia, Radioterapia AO Santa Croce e Carle, Cuneo, Italy. 4 Otorinolaringoiatria, AOU Sassari, Sassari, Italy