ABSTRACT
Various factors can affect the survival of patients with oropharyngeal cancer. We assessed the expression of protein p16INK4a, Flotillin2, epidermal growth factor receptor, and other clinicopathological features and their prognostic value for this type of cancer. We gathered patient data on demographics, clinicopathological characteristics, treatment patterns, and outcomes. Histologically and by immunochemistry staining we determined expression of prognostic factors and molecular biomarkers. The primary endpoints were overall survival (OS), disease-specific survival (DSS), and disease-free survival (DFS). Survival was assessed using the Kaplan-Meier method and Cox regression model analyses of potential prognostic parameters. After a median follow-up of 78 months, the median OS was 41 months, with an event recorded in 77.8% of patients. Median DFS was 22 months, 37 patients (51.4%) had disease relapse. The DSS survival rate was 58.3% with a median survival of 68 months. In regards to molecular biomarkers previously mentioned, there was no statistical significance for survival categories. After conducting a multivariate analysis of significant variables, we found that only recurrence, vascular invasion, and surgical intervention remained as factors with independent effects on both OS and DFS. Recurrence and the N stage were identified as independent prognostic factors for DSS. Our analysis underscores the complexity of factors that collectively influence survival following the diagnosis of OPSCC. Several factors were found to be statistically significant. These factors included the type of surgical procedure, disease relapse, vascular invasion, lymphatic invasion, perineural invasion, advanced T stage of the disease, N stage of the disease, and smoking status. The significance of these factors may vary across different types of survival. This analysis did not find any significant impact on survival from the growth factors tested, namely epidermal growth factor receptor, Flotillin2, and p16INK4a, in the applied regression models.