Construction and Validation of a Risk-Scoring Model that Preoperatively Predicts Lymph Node Metastasis in Early Gastric Cancer Patients.

ABSTRACT

OBJECTIVE: The aim of this study was to create a risk-scoring model to preoperatively predict the incidence of lymph node metastasis (LNM) in early gastric cancer (EGC) patients to guide treatment. METHODS: To construct the risk-scoring model, we retrospectively analyzed a primary cohort of 548 EGC patients. Univariate analysis and logistic regression were performed. A risk-scoring model for predicting LNM in EGC patients was developed based on preoperative factors, and another cohort of 73 patients was then analyzed to validate the model. RESULTS: In the primary cohort, LNM was pathologically confirmed in 72 (13.1%) patients. In the multivariate analysis, the presence of ulceration and tumor size on gastroscopy, undifferentiated histological type, and presence of enlarged lymph nodes on computed tomography or endoscopic ultrasonography were independent risk factors for LNM. A 17-point risk-scoring model was developed to predict LNM risk. The cut-off score of the model was 8, and the area under the receiver operating characteristic curve (AUC) of the model was 0.835 [95% confidence interval (CI) 0.784-0.886]. In the validation cohort, the AUC of the model was 0.829 (95% CI 0.699-0.959). CONCLUSIONS: We developed and validated an effective 17-point risk-scoring model that could preoperatively predict LNM for EGC patients.