The optimal technique for superficial non-ampullary duodenal lesions (SDLs) is uncertain because potential gains in resection quality with ESD may trade off against safety. Thus, this study aimed to systematically review and synthesize the evidence comparing EMR and ESD techniques for efficacy and adverse events in this population. 

We searched MEDLINE and Embase (via Ovid) to July 14, 2025, for comparative studies. Two reviewers independently screened/extracted data assessed risk of bias and graded the level of certainty (GRADE). Random-effects models yielded risk ratios (RRs) and 95%CIs. A prespecified subgroup analyzed studies where the EMR arm reported mean/median lesion diameter ≥20 mm.

Twenty-four retrospective cohort studies (5,515 lesions: EMR 3,709; ESD 1,806) were included. ESD improved en bloc (RR 1.10, 95% CI 1.05–1.16) and R0 resection (RR 1.15, 95% CI 1.01–1.31); local recurrence did not differ (RR 0.62, 95% CI 0.28–1.39). Procedural risk was higher with ESD, including intraprocedural perforation (RR 9.34, 95% CI 6.03–14.45) and delayed perforation (RR 5.82, 95% CI 3.09–10.96); delayed bleeding and the need for surgery or endoscopic reintervention were also increased. In the ≥20 mm subgroup, differences diminished and only intraprocedural perforation remained higher (RR 3.76, 95% CI 1.04–13.58) with ESD. Certainty of evidence (GRADE) was low for efficacy (en bloc, R0) and local recurrence, high for perforations, delayed bleeding, and surgical intervention, moderate for endoscopic reintervention, and low for mortality. 

For SDLs, ESD achieves better resection quality but at greater significant procedural risks, particularly in smaller lesions. EMR is a reasonable default first-line strategy for most benign polyps, reserving ESD for lesions where en bloc histology would alter management and in expert centers. Clinical trials, particularly with size-stratification, are needed.