Utility of random neosquamous and cardia biopsy specimens during surveillance after dysplastic Barrett's eradication:...

Surveillance after complete eradication of dysplastic Barrett's esophagus (BE) is important, given the risk of recurrent intestinal metaplasia and neoplasia; however, the optimal surveillance strategy remains unclear. This study aims to ascertain the yield of random biopsy specimens of the neosquamous epithelium (NE) and gastric cardia for detecting dysplasia.

In this prospective single-center study, patients undergoing postendoscopic eradication therapy surveillance for dysplastic BE were included. High-definition white-light, narrow-band, and near-focus imaging were used for esophageal assessment. Targeted biopsy sampling was performed on visible abnormalities, followed by 6 random cardia biopsy specimens and 4-quadrant NE biopsy specimens taken at 1-cm intervals.

Seventy-one patients underwent 119 surveillance endoscopies after complete eradication of intestinal metaplasia, yielding 2892 biopsy samples (66 targeted, 714 random cardia, 2112 NE). Targeted biopsy procedures detected intestinal metaplasia in 15.2% (10/66) and dysplasia in 3% (2/66) of biopsy specimens, leading to further treatment in 8 patients. In contrast, intestinal metaplasia was detected in 2.4% (17/714) of random cardia biopsy specimen and 0.4% (9/2112) of random NE biopsy specimen. No dysplasia was detected via random cardia or NE biopsy specimen. The total cost to detect 1 case of intestinal metaplasia via random biopsy specimens was $3144, whereas the cost to detect 1 case of dysplasia using the random biopsy strategy exceeded $29,673.

Targeted biopsy sampling of the NE and gastric cardia are important in detecting early BE recurrence. Additional random NE and cardia biopsy samples provide no added benefit in dysplasia detection but incur unnecessary time and cost expenditures. We recommend targeted biopsy sampling as the primary surveillance strategy after complete eradication of intestinal metaplasia and eliminating random biopsy sampling in expert centers.