Abstract
OBJECTIVE: To identify risk factors for Lymph Node Metastasis (LNM) in pT3-4 Laryngeal Squamous Cell Carcinoma (LSCC) patients with negative margins, and develop a nomogram to predict LNM risk. METHODS: 872 patients were divided into training (2010-2014) and validation (2015-2016) cohorts. Univariate and multivariate analyses identified LNM risk factors. A nomogram incorporating significant factors was developed in the training cohort. RESULTS: Smoking history, maximal tumor diameter ≥3.0 cm, depth of tumor invasion >1.0 cm, and supraglottic tumor location were significantly associated with LNM on multivariate analysis. A predictive nomogram incorporating these factors showed good discrimination (C-index > 0.7) in both cohorts. Patients were stratified into low, moderate and high-risk subgroups based on total risk scores. CONCLUSIONS: A LNM risk prediction model and risk grouping system was established, which may aid treatment selection for pT3-4 LSCC patients. The model and algorithm could help optimize neck management for this high-risk patient population.