Objective This system aims to identify high-risk subgroups with pathologically negative lymph nodes (ypN0) in locally advanced esophageal squamous cell carcinoma (ESCC), thereby providing a basis for informed adjuvant therapy decisions. Materials and methods A retrospective study included 325 patients with locally advanced ESCC. In the training cohort, multivariable logistic regression using temporal imaging features of the largest lymph node after neoadjuvant immunochemotherapy (NICT) identified independent predictors of pathologically positive lymph-node (ypN+). A predictive model was constructed and evaluated using time-dependent receiver operating characteristic curves, calibration curves, and decision-curve analysis. Significant variables with p < 0.05 from the multivariable logistic regression analysis were used to develop the temporal lymph node stratification (TLNS) score by rounding the absolute values of their β coefficients. This score was used to stratify ypN0 patients by prognostic risk. The predictive performance of the TLNS score was further validated through comparison with the clinical N (cN) stage. Results Multivariate analysis indicated that irregular lymph node borders, pre-NICT heterogeneity, and < 30% reduction in longest diameter after NICT were correlated with ypN+ (p < 0.05). The TLNS score outperformed cN staging in predicting prognosis for ypN0 patients, with higher C-indices in the training (0.66 vs 0.50) and external validation cohorts (0.72 vs 0.59). Among ypN0 patients without adjuvant therapy, those with high TLNS scores had significantly poorer prognosis than ypN+ patients receiving adjuvant therapy. Conclusion The TLNS score stratified ypN0 patients into two groups; high scores were associated with worse postoperative survival and potential benefit from adjuvant therapy. Key Points <Emphasis Type="BoldItalic">Question</Emphasis> The requirement of adjuvant therapy in patients exhibiting ypN0 following neoadjuvant immunotherapy combined with chemotherapy (NICT). <Emphasis Type="BoldItalic">Finding</Emphasis> Patients were categorized into high-TLNS (scores ≥ 2) and low-TLNS (scores < 2) groups based on TLNS score. <Emphasis Type="BoldItalic">Clinical relevance</Emphasis> TLNS score effectively stratified ypN0 patients into two distinct subgroups: those with a high TLNS score had reduced postoperative survival and could benefit from adjuvant therapy. Graphical Abstract

Temporal lymph node imaging features for prognostic risk stratification in neoadjuvant immunotherapy for esophageal cancer: a multicenter study

期刊论文
发表日期：

2026年4月27日

查看全文

Tiantian Fan,
Xinxin Wang,
Funing Chu,
Can Yu,
Tong Wei,
Qiuju Zhang,
Jinrong Qu,
Yang Zhou

摘要

Objective

This system aims to identify high-risk subgroups with pathologically negative lymph nodes (ypN0) in locally advanced esophageal squamous cell carcinoma (ESCC), thereby providing a basis for informed adjuvant therapy decisions.

Materials and methods

A retrospective study included 325 patients with locally advanced ESCC. In the training cohort, multivariable logistic regression using temporal imaging features of the largest lymph node after neoadjuvant immunochemotherapy (NICT) identified independent predictors of pathologically positive lymph-node (ypN+). A predictive model was constructed and evaluated using time-dependent receiver operating characteristic curves, calibration curves, and decision-curve analysis. Significant variables with p < 0.05 from the multivariable logistic regression analysis were used to develop the temporal lymph node stratification (TLNS) score by rounding the absolute values of their β coefficients. This score was used to stratify ypN0 patients by prognostic risk. The predictive performance of the TLNS score was further validated through comparison with the clinical N (cN) stage.

Results

Multivariate analysis indicated that irregular lymph node borders, pre-NICT heterogeneity, and < 30% reduction in longest diameter after NICT were correlated with ypN+ (p < 0.05). The TLNS score outperformed cN staging in predicting prognosis for ypN0 patients, with higher C-indices in the training (0.66 vs 0.50) and external validation cohorts (0.72 vs 0.59). Among ypN0 patients without adjuvant therapy, those with high TLNS scores had significantly poorer prognosis than ypN+ patients receiving adjuvant therapy.

Conclusion

The TLNS score stratified ypN0 patients into two groups; high scores were associated with worse postoperative survival and potential benefit from adjuvant therapy.

Key Points

Question The requirement of adjuvant therapy in patients exhibiting ypN0 following neoadjuvant immunotherapy combined with chemotherapy (NICT).

Finding Patients were categorized into high-TLNS (scores ≥ 2) and low-TLNS (scores < 2) groups based on TLNS score.

Clinical relevance TLNS score effectively stratified ypN0 patients into two distinct subgroups: those with a high TLNS score had reduced postoperative survival and could benefit from adjuvant therapy.

Graphical Abstract