Development and validation of an explainable model for predicting lymph node metastasis in early gastric…

Cohort studyEuropean journal of surgical oncology : the journal of the European Society of Surgical Oncology and the British Association of Surgical Oncology· Oct 2026

Development and validation of an explainable model for predicting lymph node metastasis in early gastric cancer: a multicenter retrospective study with external and prospective validation.

Ran Xiong, Xiangfei Sun, Qi Jiang … Xiaodong Gao

Read paper

Prediction Models & Risk ScoresCohort studyPrognosis

XGBoost model accurately predicts lymph node metastasis in early gastric cancer.

Development and validation of an explainable model for predicting lymph node metastasis in early gastric cancer: a multicenter retrospective study with external and prospective validation.

Ran Xiong … Xiaodong Gao

European journal of surgical oncology · Oct 2026

Background

Accurate preoperative prediction of lymph node metastasis (lymph node metastasis) is essential for treatment selection in early gastric cancer (early gastric cancer), particularly when considering endoscopic submucosal dissection.

Purpose

We aimed to develop and validate an interpretable machine-learning model for individualized preoperative prediction of lymph node metastasis in early gastric cancer.

Methods

With pathologically confirmed early gastric cancer who underwent gastrectomy with lymph node dissection were retrospectively enrolled for training and external validation.

n = 3412 patients

Results

model accurately predicted lymph node metastasis (AUC; 1.0 is perfect) in prospective validation

training AUC

0.9289

external AUC

0.8583

prospective AUC

0.8766

preoperative pathology

0.8653

More results

Among 3412 patients in the training cohort, 467 (13.7%) had lymph node metastasis.

In the preoperative pathology validation cohort, the model achieved an AUC of 0.8653 and a PR-AUC of 0.6050.

Key predictors included lymphovascular/neural invasion, invasion depth, tumor location, differentiation, and tumor size.

“

Conclusion

This externally and prospectively validated, interpretable machine-learning model provides transparent individualized risk estimates and may support personalized decision-making between endoscopic and surgical management in early gastric cancer.

Development and validation of an explainable model for predicting lymph node metastasis in early gastric cancer: a multicenter retrospective study with external and prospective validation.

Development and validation of an explainable model for predicting lymph node metastasis in early gastric cancer: a multicenter retrospective study with external and prospective validation.

Related papers