Development and validation of a nomogram for prediction of cervical lymph node metastasis in middle and lower thoracic esophageal squamous cell carcinoma.

Background: Estimates of cervical lymph node (LN) metastasis in patients with middle and lower thoracic esophageal squamous cell carcinoma (ESCC) are important. A nomogram is a useful tool for individualized prediction.

Methods: A total of 235 patients were enrolled in this study. Univariate and multivariate analyses were performed to screen for independent risk factors and construct a nomogram to predict the risk of cervical LN metastasis. The nomogram performance was assessed by discrimination, calibration, and clinical use.

Results: Totally, four independent predictors, including the maximum diameter of tumor, paraesophageal lymph node status, recurrent laryngeal nerve lymph node status, and the CT-reported cervical LN status, were enrolled in the nomogram. The AUC of the nomogram model in the training and validation dataset were 0.833 (95% CI 0.762-0.905), 0.808 (95% CI 0.696-0.920), respectively. The calibration curve demonstrated a strong consistency between nomogram and clinical findings in predicting cervical LN metastasis. Decision curve analysis demonstrated that the nomogram was clinically useful.

Conclusion: We developed a nomogram that could be conveniently used to predict the individualized risk of cervical LN metastasis in patients with middle and lower thoracic ESCC.