Polyps in the gastrointestinal tract: discrepancy between endoscopic forceps biopsies and resected specimens.

Background And Aim: An endoscopic forceps biopsy (EFB) carries the risk of missing the neoplastic foci within a polyp as only a small portion of the lesion is sampled using this technique. Accordingly, the histological examination of specimens obtained by an EFB is of limited accuracy and patient management based on the histological findings is controversial. The aim of this prospective study was to determine the diagnostic concordance between an EFB and resected tissues of gastric and colon polyps.

Methods: Between January 2006 and October 2007, 1312 gastrointestinal polyps from 896 patients were resected in our hospital. Patients with polyps of epithelial origin of at least 8 mm in diameter and not associated with polyposis syndromes were included in the study. Polyps of nonepithelial origin were excluded. One thousand two hundred and sixty-four polyps of epithelial origin [gastric polyps (n=268) and colon polyps (n=996)] obtained from 813 patients met the inclusion criteria. All patients underwent an EFB and resection of the polyp by endoscopic mucosal resection and endoscopic submucosal dissection.

Results: Multiple polyps existed in 31.6% of the patients. The pathological diagnoses of resected gastric polyps were as follows: adenomas with low-grade dysplasia, 46 (17.2%); adenomas with high-grade dysplasia, 42 (15.7%); hyperplastic polyps, 126 (47.0%); chronic inflammatory polyps, 29 (10.8%); and adenocarcinomas, 25 (9.3%). The discrepancy rate between an EFB and the pathology of the resected gastric polyps was 39.2% (the Kendall's tau-b and the kappa coefficient for agreement between the EFB and resected specimens of gastric polyps were 0.577 and 0.472, respectively; P value <0.001). No relationship between the size of the gastric polyp and the concordance rate was observed. The pathological diagnoses of the resected colon polyps were as follows: adenomas with low-grade dysplasia, 559 (56.1%); adenomas with high-grade dysplasia, 229 (23.0%); hyperplastic polyps, 44 (4.4%); adenocarcinomas, 53 (5.3%); and inflammatory polyps, 111 (11.1%). The discrepancy rate between the EFB and the pathology of the resected colon polyps was 39.8%. (the Kendall's tau-b and the kappa coefficient for agreement between the EFB and the resected specimens of the colon polyps were 0.479 and 0.293, respectively; P value <0.001). No relationship between the size of the colon polyp and the concordance rate was observed.

Conclusion: Considerable discrepancies were observed in histological findings between the EFB and the resected specimens. Therefore, complete removal of the entire polyp is recommended to confirm the diagnosis, to remove precancerous lesions, and to develop an optimal management plan.