Long-Term Outcome of Endoscopic Resection vs. Surgery for Early Gastric Cancer: A Non-inferiority-Matched Cohort Study

Abstract

OBJECTIVES: 

Few studies have compared the long-term outcomes of endoscopic resection and surgery. The aim of this study was to compare the long-term outcomes of endoscopic resection with those of surgery for early gastric cancer (EGC).

METHODS: 

We reviewed prospectively collected data of patients who had undergone endoscopic resection (1,290 patients) or surgery (1,273 patients) for EGC. To reduce the effect of selection bias, we performed a propensity score-matching analysis between the two groups. The primary outcome was overall survival (OS). The secondary outcomes were disease-specific survival, disease-free survival (DFS), recurrence-free survival (RFS), occurrence of metachronous gastric cancer, treatment-related complications, length of hospital stay, and 30-day outcomes. The study was designed as a non-inferiority study and tested in an intention-to-treat analysis.

RESULTS: 

In a propensity-matched analysis of 611 pairs, the 10-year OS proportion was 96.7% in the endoscopic resection group and 94.9% in the surgery group (P=0.120) (risk difference −1.8%, 95% confidence interval (CI) −4.04–0.44,P_{non-inferiority}=0.014), which met the non-inferiority criterion. In contrast, the 10-year RFS proportion was 93.5% in the endoscopic resection group and 98.2% in the surgery group (P<0.001) (risk difference 4.7%, 95% CI 2.50–6.97,P_{non-inferiority}=0.820), which did not meet the non-inferiority criterion, mainly because of metachronous recurrence in the endoscopic resection group. The rate of early complications was higher in the endoscopic resection group than in the surgery group (9.0 vs. 6.6%,P=0.024), whereas the rate of late complications was higher in the surgery group than in the endoscopic resection group (0.5 vs. 2.9%,P<0.001). In the multiple Cox regression analysis, patient’s age, the comorbidity index, the performance index, sex, tumor morphology, and depth of invasion were predictors of OS in patients with EGC.

CONCLUSIONS: 

Endoscopic resection might not be inferior to surgery with respect to OS in patients with EGC lesions that meet the absolute or expanded criteria. However, DFS, RFS, and metachronous RFS might be lower after endoscopic resection than after surgery.