Head and neck squamous cell carcinoma (HNSCC) is a challenging cancer associated with significant morbidity and mortality. By exploring diagnostic, prognostic, and therapeutic strategies, the author aimed to improve the clinical management of this complex disease.

First, we identified biomarkers predictive of occult lymph node metastases (OLNM), a significant determinant of prognosis in HNSCC. We developed a Random Forest model by combining clinical, pathological, and molecular data, achieving high diagnostic accuracy in predicting OLNM. This underscores the transformative potential of artificial intelligence (AI) in synthesizing complex datasets and informing clinical decisions. We also evaluated advanced imaging techniques, particularly FDG PET-CT, for OLNM detection. Although FDG PET-CT demonstrated high specificity and NPV, its limited sensitivity suggests the need for complementary diagnostic methods to enhance accuracy.

During the author’s MD-PhD thesis, he developed an animal model to study postsurgical recurrences of HNSCC, which faithfully recapitulated key aspects of human disease, including local tumor recurrence and regional lymphatic spread. This model allowed further study of HNSCC biology, notably FOXM1, a key transcription factor in the invasion of head and neck cancer.

This Privat Docent (PD) thesis represents the results of the author’s research that focused on the problem of OLNM and disease progression with the help of preclinical models, thereby laying a strong foundation for improving patient outcomes. Continued research based on these insights holds promise for addressing the unmet needs of HNSCC patients and advancing the field of head and neck oncology.