Related Experiment Video
Updated: Aug 27, 2026
Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
Comparison and validation of two proposed modifications to the current lymph node classification for esophageal
Li Fu1, Ligong Yuan2, Donghui Jin3
1Department of Thoracic Surgery, National Cancer Center/National Clinical Research Center for Cancer/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College, Beijing, 100021, China.
Background:
Accurate lymph node (LN) classification is crucial for prognostic evaluation and individualized management in esophageal squamous cell carcinoma (ESCC).
Methods:
We retrospectively analyzed 688 ESCC patients who underwent R0 resection without neoadjuvant therapy, dividing them into development (n = 461) and validation (n = 227) cohorts. The prognostic performance of Lymph Node Ratio (LNR) and Log Odds of Positive Lymph Nodes (LODDS) was compared against the AJCC N-staging system. Multivariable Cox regression models were used to identify independent predictors and construct nomograms for Overall Survival (OS) and Recurrence-Free Survival (RFS). Model performance was assessed via the C-index, time-dependent ROC curves, and Decision Curve Analysis (DCA).
Results:
Multivariable analysis identified LODDS as a modestly improved prognostic factor compared to AJCC-N and LNR, exhibiting slightly improved discriminatory performance for OS (C-index: 0.687; Likelihood Ratio: 102.9) and RFS (C-index: 0.685; Likelihood Ratio: 107.8). Based on these findings, LODDS-incorporated nomograms were constructed. In the validation cohort, these nomograms displayed robust predictive accuracy, with 1-, 3-, and 5-year AUCs of 0.737, 0.765, and 0.788 for OS, and 0.689, 0.748, and 0.783 for RFS, respectively. Calibration plots showed excellent agreement between predicted and observed survival, and DCA confirmed that the LODDS-based nomograms provided greater net clinical benefit than traditional staging systems.
Conclusion:
Among the three staging systems evaluated, LODDS provided complementary prognostic information beyond conventional lymph node classifications and may improve postoperative risk stratification in ESCC patients. The proposed LODDS-based nomograms demonstrate robust predictive accuracy and clinical utility, serving as valuable tools for individualized postoperative management.
