Background and study aims: Endoscopic mucosal resection (EMR) has been carried out for high-grade dysplasia (HGD) and intramucosal carcinoma (IMCA) in Barrett's esophagus using two different cap-assisted techniques, the "inject, suck, and cut" and the "band and snare." Previous work has demonstrated comparable specimen diameters. However, the two techniques have not been previously compared with respect to depth of resection.
Patients and methods: From a database of patients with Barrett's esophagus, we identified 40 consecutive specimens removed using EMR from patients with HGD or IMCA, 20 each from the "inject, suck, and cut" and the "band and snare" techniques. Specimens were evaluated and measured separately by two pathologists for greatest diameter and depth, and for the presence of submucosa and muscularis propria at the deepest margin of resection. Follow-up data were collected regarding clinical outcome and stricture formation.
Results: The mean depth of the specimens from the two techniques was not significantly different (0.51 cm vs. 0.50 cm, P = 0.76). All specimens contained substantial submucosa, allowing accurate staging of the neoplastic lesions resected. Muscularis propria was identified at the base of 65% of the "band and snare" and 50% of the "inject, suck, and cut" specimens (P = 0.52).
Conclusions: The "inject, suck, and cut" and "band and snare" techniques both yield equivalent adequate depth of histological specimens from Barrett's esophagus with HGD or IMCA, and both provide accurate pathological staging.