Duodenal endoscopic submucosal dissection (ESD) is historically considered one of the highest-risk endoscopic resections in Western practice due to thin wall anatomy and high perforation rates reported in earlier studies. As a result, Western adoption remains limited. We aimed to evaluate the feasibility, safety, and oncologic performance of duodenal ESD compared with EMR over a 5-year period in a Western expert center.

We retrospectively analyzed all non-ampullary duodenal lesions resected between 2019–2024. Patients were treated with EMR (n=8) or ESD (n=7). Demographics (age, sex, comorbidities), lesion size, D1–D3 location, morphology, and Vienna classification were recorded. Mean lesion size was comparable between EMR and ESD (31.0 vs 26.1 mm; p=0.77). Sex distribution (62.5% vs 57.1% male) and comorbidities (62.5% vs 42.8%; p=0.62) were similar, indicating no selection bias. Paris classification included predominantly superficial elevated (0-IIa) and flat mixed patterns (0-IIa/Is). Statistical analysis included Fisher’s exact test and Mann–Whitney U. Correlation between lesion size and en-bloc resection was assessed via Spearman’s coefficient. Procedural time could not be analyzed due to incomplete documentation.

ESD achieved 100% en-bloc resection (7/7) compared with 42.8% for EMR (p=0.028). R0 resection rates were identical (83.3%) when margins were assessable.Safety outcomes were excellent: no perforations (0%), no delayed bleeding, and no patient required surgery. One immediate intraprocedural bleed occurred in the ESD group (14.2%), successfully controlled endoscopically.

Critically, most ESDs were performed in D2 (6/7), the anatomically highest-risk duodenal segment. ESD achieved a 100% en-bloc rate in D2, with zero perforations and zero delayed bleeding, demonstrating exceptional safety even in the most challenging location.

Paris classification contributed to meaningful selection differences. EMR was used mainly for superficial elevated 0-IIa or Is lesions suited for lifting–snare resection. ESD was preferentially used for flat or mixed 0-IIa/0-IIa+Is lesions, which traditionally carry higher risks of incomplete resection with standard EMR.Despite this, ESD maintained 100% en-bloc across all Paris morphologies.

Lesion size did not correlate with en-bloc success (Spearman ρ = –0.26, p=0.34), indicating that ESD maintained stable technical performance across increasing lesion dimensions. Recurrence occurred in one ESD patient, while none was observed after EMR.

Patient factors (age, sex, comorbidities), lesion size, and anatomical location did not influence technique selection or outcomes, supporting the robustness of ESD performance in this cohort.

In this 5-year Western cohort, duodenal ESD demonstrated perfect en-bloc performance and zero perforations, overturning long-standing concerns about duodenal ESD safety in Western practice. The ability to achieve 100% en-bloc resection even in D2—while maintaining a 0% perforation rate—represents a major technical achievement and aligns with high-quality Asian data.ESD should be strongly considered for large or complex non-ampullary duodenal lesions requiring definitive histologic assessment. EMR remains appropriate for small, low-risk lesions; however, for curative-intent resections, ESD offers clearly superior technical and oncologic performance.