Endoscopic Mucosal Resection versus Endoscopic Submucosal Dissection for Barrett's Neoplasia: Clinical Outcomes at a Canadian Tertiary Referral Center

Aims Endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD) are established techniques for endoscopic resection of Barrett’s neoplasia, but consensus is lacking for which offers superior outcomes. This study compares the outcomes of EMR and ESD at a Canadian tertiary Barrett’s referral center.

Methods A retrospective single-center study at St. Michael’s Hospital, Toronto, included patients treated for Barrett’s neoplasia (high-grade dysplasia or esophageal adenocarcinoma) from January 2015 to August 2023. Primary outcome was local recurrence after complete remission of dysplasia (CRD), with secondary outcomes being persistent dysplasia at the first follow-up and CRD achievement. Follow-up was performed until November 2023.

Results The study included 331 patients: 72 underwent ESD and 259 underwent EMR (mean age: 68.5 y/o, 90.3% male in ESD, 67.6 y/o, 80.3% male in EMR). ESD group had larger lesions (4.5 vs 3.0cm, P<0.05) and higher circumferential occupancy (53.4% vs 43.0%, P<0.05). Adverse events were slightly more common in ESD (bleeding: 2.8% vs 0.4%, perforation: 1.4% vs 0.8%, P=0.15). Histologically, adenocarcinoma was more prevalent in ESD (94.4% vs 77.2%, P<0.05), with higher incidence of deep submucosal invasion (M: 75% vs 82% SM1: 6% vs13%, SM2: 19% vs 6%, P<0.05). Curative resection rates were 63.9% for ESD and 87.3% for EMR (defined as the absence of high-risk pathologic features for lymph node metastasis regardless of the vertical margin). Of the 275 patients who underwent follow-up (54 ESD, 221 EMR), persistent dysplasia at the first follow-up was significantly lower in ESD (11.1% vs 25.8%, P<0.05). The total number of resections until CRD was similar (1.3 vs. 1.8 times, P=0.14). CRD was achieved in 100% of ESD patients and 96.8% (214/221) of EMR patients (P=0.35). For patients who didn’t achieve CRD after EMR, 6/7 underwent esophagectomy and 1/7 entered palliative care. Local recurrence after CRD was significantly lower in ESD (1.9% [1/54] vs 12.5% [26/214], P<0.05). Local recurrence after ESD was successfully treated with endoscopy, while in the EMR group, 21/26 were successfully treated with endoscopy, 3/26 required esophagectomy, and 2/26 entered palliative care. The median total follow-up duration was significantly longer in EMR group [288 (130-524) days vs 623 (318-1171) days, P<0.05].

Conclusions Despite non-randomized treatment selection in which more advanced or concerning appearing lesions were selected for ESD, these patients showed significantly lower rates of persistent dysplasia at the initial follow-up. Furthermore, after achieving CRD, the ESD group demonstrated a reduced incidence of local recurrence. In contrast, some local recurrences after EMR necessitated esophagectomy, with some cases leading to palliative care. These findings suggest that ESD may provide a more favorable prognosis in the treatment of Barrett’s esophagus compared to EMR.

Conflicts of interest

JDM – Speaker: Boston Scientific, Pendopharm, SCOPE rounds, Vantage, Medtronic. Medical Advisory Board: Pendopharm, Boston Scientific, Janssen, Pentax, Fuji. Grants and Research support: CAG.GRM – Consultant for Olympus. Speaker: Pentax, Fuji and Medtronic.CWT – Speaker: Medtronic and Boston Scientific, Consultant: Boston Scientific.

Publication History

Article published online:
15 April 2024