RESUMO
BACKGROUND: This study elucidates the clinical impact of surgical treatment of head and neck squamous cell carcinoma (HNSCC) based on a detailed search of all exons of the TP53 gene and p53 protein phenotypic analysis using formalin-fixed paraffin-embedded (FFPE) specimens. METHODS: Clinically well-annotated FFPE specimens from 317 patients with HNSCC treated by surgery were examined by all-exon TP53 sequencing using a next-generation sequencer and p53 protein phenotype by immunohistochemistry. After excluding human papillomavirus-associated oropharyngeal carcinomas, two risk categories were classified as "p53 adverse function" and "p53 favorable function" based on TP53 mutation status and p53 protein phenotype. Mutation in PIK3CA, AKT, and HRAS was also evaluated by target sequence. Cox proportional hazards regression models were used for statistical analysis of clinical outcomes. Receiver operating characteristic curve analysis was used to determine the optimal surgical margin cutoff for local recurrence. Local control rates were compared between the risk groups using Fisher's exact test. RESULTS: Multivariate analysis identified "p53 adverse function" as an independent poor predictor of overall survival, local control, and distant metastasis-free survival. In oral cavity cancer, the optimal surgical margin cutoff associated with local recurrence was 6 mm. In patients with surgical margin > 6 mm, the "p53 adverse function" group demonstrated significantly higher local recurrence rate than the "p53 favorable function" group. PIK3CA, AKT, or HRAS mutation did not correlate with improved overall survival. CONCLUSIONS: All-exon TP53 sequencing and p53 protein phenotype analysis using FFPE specimens can accurately predict clinical outcomes.