A Meta-analysis of Endoscopic Submucosal Dissection and EMR for Early Gastric Cancer

Background: Endoscopic submucosal dissection (ESD) was developed to overcome the problem caused by incomplete resection by conventional EMR for early gastric cancer. However, the technique of ESD had a relative higher incidence rate of complications such as bleeding and perforation because of its large wound incidence and difficulties.

Objective: To evaluate the efficacy and safety of ESD and EMR for early gastric cancer.

Design: We made an overall strategy to search various databases both electronically and manually and assessed the quality of selected articles according to the inclusion and exclusion standard.

Setting: Meta-analysis of 9 retrospective studies.

Patients: 3548 lesions (ESD 1495; EMR 2053).

Intervention: ESD or EMR.

Main Outcome Measurements: En bloc resection rate, histologic curative resection rate, complications, recurrence rate.

Results: Nine retrospective studies were identified (7 full text and 2 abstracts). A total of 3548 lesions were enrolled (ESD 1495; EMR 2053). The mean time required for resection was longer for ESD than for EMR (weighted mean difference 59.4; 95% confidence interval [CI], 16.8-102.0); the en bloc rate in the ESD group was significantly higher than that in the EMR group (OR 9.69; 95% CI, 7.74-12.13), and so was the total histologically complete resection rate (OR 5.66; 95% CI, 2.92-10.96). The ESD group had lower recurrence frequency (OR 0.10; 95% CI, 0.06-0.18). The perforation rate was higher in the ESD group (OR 4.67; 95% CI, 2.77-7.87), whereas the bleeding incidences were similar between the two groups.

Limitations: Heterogeneity was present among the studies.

Conclusion: Compared with EMR for early gastric cancer, ESD showed considerable advantages regarding en bloc resection rate, histologically complete resection rate, and local recurrence even for small lesions, but it had the disadvantages of higher complication rates for perforation. All of the results mentioned should be confirmed by well-designed, randomized, controlled trials from more countries, with larger samples and long enough follow-up periods.