Is the risk of concomitant invasive esophageal cancer in high-grade dysplasia in Barrett's esophagus overestimated?

Background & Aims: Recent studies have claimed long neoplasia-free survival rates with endoscopic mucosal resection of high-grade dysplasia (HGD) in Barrett's esophagus (BE). However, reports have contended that approximately 40% of patients who have esophagectomy for HGD have occult invasive cancer. The aim of this study was to use explicit criteria to determine the true prevalence of invasive adenocarcinoma in reports of patients who had esophagectomy for HGD in BE. Methods: Studies reporting rates of esophageal cancer in patients who underwent esophagectomy for HGD in BE were gathered using MEDLINE and PUBMED. We defined invasive esophageal adenocarcinoma (IEAC) as tumor with submucosal invasion or beyond. Intramucosal carcinoma (IMC) was not considered IEAC. Results: Twenty-three articles were selected for analysis. Most investigators reported rates of invasive cancer in the esophagectomy specimen, and the pooled average was 39.9% among the 441 patients who underwent an esophagectomy for HGD. Reported rates varied from 0% to 73%. A total of 267 patients had American joint Committee on Cancer stage 0 postoperatively, 132 patients had stage 1, 23 patients had stage IIa, 10 patients had stage IIb, and 9 patients had stage III. Fourteen studies provided differentiation between intramucosal and submucosal. invasion. Among 213 patients, only 12.7% had IEAC, whereas 87.3% had HGD or IMC. The IEAC rate of 11% among patients with visible lesions is greater than the rate of 3% among patients with no visible lesion. Conclusions: By using strict pathologic definitions of invasive disease, the present study indicates the true prevalence of IEAC in BE and HGD may have been overestimated significantly. Separating IMC from IEAC is clinically relevant because endoscopic techniques potentially may treat IMC.