Objective To compare the long-term prognosis differences between surgical radical resection and endoscopic resection for early gastric cancer patients based on the SEER database.

Methods 1 437 patients with stage Tis to T₁b gastric adenocarcinoma were selected from the SEER database from January 1, 2004 to December 31, 2013. They were divided into a surgical group (n=1 257) and an endoscopic resection group (n=180) according to the treatment method. Kaplan Meier survival curve and Cox regression model were used to analyze survival outcomes.

Results The patients in the surgery group were younger than those in the endoscope group (67.63 vs 71.29 years old), with higher rates of T₁ stage (97.45% vs 87.78%) and lymph node metastasis (19.73% vs 5.00%; all P＜0.001). The median follow-up time for all patients was 37 (15, 66) months, and the mortality rate of gastric cancer in the endoscopic group was lower than that in the surgical group (23.33% vs 27.13%, P＜0.001). Univariate Cox analysis showed that treatment modality, age, sex, T stage, lymph node metastasis are all associated with early gastric cancer mortality(all P＜0.05), and the risk of death in the endoscope group is 43% of that in the surgery group(HR=0.43, P=0.015). After adjusting for multiple factors, there was no statistically significant difference in mortality risk between the two groups (P=0.067), but after excluding lymph node positive patients, the mortality risk in the endoscope group was 46% of that in the surgery group (HR=0.46, P=0.048).

Conclusions For early gastric cancer patients with negative lymph nodes, endoscopic resection may provide better survival benefits than surgical procedures, suggesting that it can be the preferred treatment strategy for patients with low risk of lymph node metastasis.