Prediction model for cervical lymph node metastasis using CT characteristics in patients with head and neck squamous cell carcinoma.

Abstract

e17547 Background: Lymph node metastasis is a major prognostic determinant for patients with head and neck squamous cell cancer (HNSCC). This study aimed to construct a predictive model using CT characteristics of lymph node and tumor for patients with HNSCC to stratify the risk of lymph node metastasis. Methods: The study population was obtained from historical cohort of 472 cervical lymph nodes from 191 patients with HNSCC in a tertiary referral hospital. We analyzed preoperative CT images of lymph nodes according to diameter, ratio of long-to-short axis diameter, necrosis or cystic change, conglomeration, infiltration to adjacent soft tissue, and laterality (ipsilateral vs contralateral) and analyzed T stage. Reference standard was surgical pathologic results. Multivariate logistic regression analysis was performed to predict whether nodules were diagnosed as metastasis or benign. Results: CT features of lymph nodes, including minimal axial diameter, ratio of long-to-short axis diameter, necrosis or cystic change, and T stage were selected as predictors for lymph nodes metastasis. A 10-point risk scoring system was developed, and the risk of malignancy ranged from 7.3% to 99.8%, which was positively associated with increases in risk scores. The areas under the receiver operating characteristic curve of the development and validation sets were 0.886 and 0.879, respectively. Conclusions: We have devised a simple predictive model using the CT characteristics of lymph nodes and tumor for HNSCC to stratify the risk of cervical lymph node metastasis. [Table: see text]