Abstract
BACKGROUND: The prognostic role of preoperative inflammation-nutrition-related markers in locally advanced lip squamous cell carcinoma (LSCC) remains underexplored. This study aimed to assess the impact of various preoperative inflammation-nutrition-related markers on the prognosis of patients with locally advanced LSCC undergoing surgical treatment and to establish a corresponding predictive model. METHODS: A retrospective analysis was performed on the clinical data of 169 patients with locally advanced LSCC who underwent surgical treatment. A total of 27 clinicopathological variables, including inflammation-nutrition-related markers, were collected. Univariate and multivariate Cox regression analyses were used to identify independent prognostic factors for disease-free survival (DFS) and overall survival (OS). The nomogram models were validated using receiver operating characteristic (ROC) curve analysis, calibration plots, and decision curve analysis (DCA). Risk stratification was performed based on the nomogram scores, and differences between risk subgroups were explored. RESULTS: The extranodal extension (ENE), surgical safety margin, Glasgow prognostic score (GPS), Geriatric Nutritional Risk Index (GNRI), Controlling Nutrition score (CONUT), American Joint Committee on Cancer (AJCC) stage, and adjuvant radiotherapy were independent prognostic factors for DFS. In contrast, ENE, surgical safety margin, GNRI, CONUT, AJCC stage, and adjuvant radiotherapy were also independent prognostic factors for OS. The nomograms demonstrated better predictive performance than the AJCC staging system. Based on the nomogram model, patients were stratified into low-, medium-, and high-risk subgroups, which exhibited significant differences in survival outcomes. CONCLUSION: GPS, GNRI, and CONUT are independent factors affecting the prognosis of patients with locally advanced LSCC undergoing radical surgery. By combining GPS, GNRI, and COUNT with other independent clinicopathological prognostic factors, a reliable nomogram model can be established to accurately predict patients' DFS and OS. This provides a powerful tool for individualized prognostic assessment, optimized risk stratification, and treatment decision-making.