Large population-based study using the SEER database: is endoscopic resection appropriate for early gastric cancer patients in the United States?

Abstract

Aim

To investigate whether endoscopic treatment is applicable to American patients and explores the predictors of lymph node metastasis (LNM) in early gastric cancer (EGC).

Methods

Patients with EGC confined to either mucosa (T1a, n = 1799) and submucosa (T1b, n = 1689) were identified from the Surveillance, Epidemiology, and End Result database. Multivariate logistic regression, Kaplan–Meier method, and univariate/multivariate Cox regression were used to assess the correlation between invasion depth and LNM or prognosis. A nomogram for predicting LNM was constructed and internally validated.

Results

EGC limited to T1a exhibited a 2.4% incidence of LNM, which increased to 11.1% when the depth invaded T1b.LNM was present at 1.4%, 5.2%, and 5.0% for sizes ≤2, 2–5, and >5cm of low-grade T1a EGC, respectively, (p = .019) and at 4.8%, 12.4%, and 28.6% of T1b EGC, respectively (p  < .001).The multivariate logistic model revealed that older age, T1b invasion, larger tumor size, and high-grade lesions were associated with a higher risk of LNM. Moreover, the T1a EGC patients had better cancer-specific survival (OS) and overall survival(CSS) compared with the T1b EGC patients (5-year OS: 77.2% versus 67.4%, p < .001; 5-year CSS: 90.6% versus 81.4%, p < .001). The discrimination of the prediction model was 0.745.

Conclusions

Endoscopic treatment may only be suitable for patients in the US population who have low-grade T1a lesions of less than 2 cm in size. Patients with T1a lesions of greater than 2 cm in size, lesions with high-grade, and all T1b lesions may benefit from radical surgical resection with lymphadenectomy.

Keywords:

• Early gastric cancer
• cancer-specific survival
• overall survival
• lymph node metastasis
• surveillance
• epidemiology
• and End Result database

Acknowledgments

The authors thank the Surveillance, Epidemiology, and End Results (SEER) database which provided data for our research.

Disclosure statement

The authors indicate no potential conflicts of interest.

Additional information

Funding

This work was supported by Natural Science Foundation of Tianjin (No. 18JCZDJC45200), Postgraduate Innovation Fund of Tianjin Medical University (No. YJSCX201901), Science and Technology Program of Tianjin (No. 19PTZWHZ00090), and Tianjin Research Innovation Project for Postgraduate Students (No. 2019YJSB108).