Endoscopic mucosal resection: not only therapeutic, but a diagnostic procedure for sessile gastric polyps

Abstract

Background and aims: Histological examination of specimens obtained by forceps biopsy sampling of gastric lesions is of limited accuracy, and their management on this basis is therefore controversial. Endoscopic mucosal resection (EMR) was initially developed in Japan for the resection of early gastric cancer (EGC). The potential use of EMR as a diagnostic tool has been suggested. The aims of the present study were to assess the value of forceps biopsy sampling in establishing the correct diagnosis revealed by EMR and to evaluate the efficacy of EMR.

Methods: Fifty-six subjects with sessile gastric polyps of epithelial origin, at least 0.5 cm in diameter, and not associated with polyposis syndromes, were included. Following forceps biopsy sampling, EMR was performed with an inject-and-cut technique or with cap-fitted methods. The histological results on the forceps biopsy and the resected specimens were analyzed.

Results: Histology on the resected specimens revealed neoplastic lesions in 34 cases, including seven EGC, and there were hyperplastic-inflammatory lesions in 21 cases. Complete agreement between the previous histological results of the forceps biopsy samples and the resected specimens was seen in only 76.7% of the lesions. Altogether, the sensitivity and specificity of the forceps biopsy procedure for diagnosing neoplastic lesions were 87.5% (95% confidence interval [CI] = 76.0-98.9%) and 65.2% (95% CI = 45.7-84.7), respectively. A clinically relevant discrimination between neoplastic and non-neoplastic lesions was not achieved in seven cases. No complications, such as perforation or massive bleeding necessitating surgical treatment, were encountered. EMR was considered complete in five patients. None of the EGC recurred during the mean 38-month (6-72) follow up.

Conclusions: Forceps biopsy is not fully representative of the entire lesion, and a simple biopsy may therefore lead to a faulty differentiation between neoplastic and non-neoplastic lesions. EMR proposes diagnostic and staging advantage in assessing patients with EGC as compared to forceps biopsy, because it provides more intact mucosa and submucosa for histological analysis. Sessile gastric polyps should be fully resected by EMR for a final diagnosis and (depending on the lesion size and type) possibly definitive treatment.