Lymph Node Metastasis Frequency and Development of Predictive Model in T1b Gastric Cancer: A Single-Center Study.

BACKGROUND

The optimal management of T1b gastric cancer (GC) remains controversial, largely due to the risk of lymph node metastasis (LNM). This study aimed to clarify the anatomical distribution of LNM and develop a predictive nomogram for patients with T1b GC.

PATIENTS AND

METHODS

Clinical and pathological data from 1041 patients with pathologically confirmed T1b GC who underwent D2 gastrectomy at a single high-volume center in China were retrospectively analyzed. The frequency and distribution of LNM, including skip metastases, were assessed. Predictive factors identified by logistic regression were incorporated into a nomogram, which was evaluated using receiver operating characteristic (ROC) curves, decision curve analysis (DCA), and calibration plots.

RESULTS

The overall incidence of LNM was 28.34%, with first- and second-tier rates of 25.4% and 10.3%, respectively. LNM occurred across almost all nodal stations without a consistent anatomical pattern. Lymph node skip metastasis was detected in 2.98% of patients, with no reliable predictors. Importantly, no LNM was observed in well-differentiated tumors ≤ 3 cm, and in moderately to well-differentiated tumors of this size, metastases were confined to first-tier nodes. Multivariable analysis identified female, younger age, larger tumor size, horizontal location, poor differentiation, signet ring cell component, and lymphovascular invasion as independent predictors. The nomogram incorporating these factors demonstrated good discrimination (AUC 0.83; 95% CI 0.80-0.86) and favorable calibration and clinical utility.

CONCLUSIONS

LNM in T1b GC shows a heterogeneous distribution and cannot be predicted by anatomical site alone. Sentinel lymph node biopsy with limited resection may be feasible for carefully selected patients with ≤ 3 cm, moderately to well-differentiated tumors, while D2 gastrectomy remains standard for others. The proposed nomogram may facilitate individualized surgical decision-making but requires external validation.