A nomogram to predict long-term survival after resection for esophageal cancer: An observational study in northeast China.

Abstract

This study aims to create a prognostic nomogram by combining clinicopathologic variables that are linked to the overall survival following the surgical removal of esophageal squamous cell carcinoma. A total of 224 patients with esophageal cancer who underwent surgical R0 resection were included. The construction of the nomogram involved using a multivariable Cox proportional hazards regression model. To evaluate the model's effectiveness, Kaplan-Meier curves and calibration plots were used for discrimination and calibration, respectively. Nearly half of the patients were >60 years old (45.1%), and 95.5% of the patients were male. After esophageal cancer resection, 35.7% of patients experienced complications, with 23.7% developing anastomotic stenosis and 4.5% developing a fistula. Using the backward selection of clinically relevant variables, we found that tumor located in middle thoracic (hazard ratio 2.299, 95% confidence interval 1.008-5.244), anastomotic fistula (3.028, 1.436-6.384), and vascular invasion (2.175, 1.496-3.108) were independently associated with mortality (all P < .05), whereas lymph node clearance ≥15 nodes is associated with longer survival (0.444, 0.278-0.710) (P = .001). On the basis of these factors, a nomogram was created to predict survival of esophageal squamous cell carcinoma after resection. Discrimination using Kaplan-Meier curves, calibration curves, and bootstrap cross-validation revealed good predictive abilities (C index, 0.673). A nomogram was created based on the experience from northeast China to forecast overall survival following resection for esophageal squamous cell carcinoma. The validation process demonstrated accurate distinction and calibration, indicating the practical value of the nomogram in enhancing personalized survival predictions for patients who undergo esophageal squamous cell carcinoma resection in this study population.