Background & aims: The identification of any high-grade dysplasia (HGD) in Barrett's esophagus has been considered to be an indication for esophagectomy because of the increased risk of cancer. The aim of this study was to determine if a limited extent of HGD has the same potential for cancer as diffuse HGD.
Methods: A retrospective cohort study was performed to assess the risk of developing adenocarcinoma in relationship to the extent of HGD found on endoscopic surveillance. The extent of HGD was defined as focal if cytologic and/or architectural changes of HGD were limited to a single focus of 5 or fewer crypts and diffuse if more than 5 crypts were involved in a single biopsy specimen or if HGD involved more than one biopsy fragment. The relative risk of cancer was assessed using a Cox proportional hazard model, and cancer-free survival was determined using survival curves.
Results: Sixty-seven patients with diffuse HGD and 33 with focal HGD satisfied selection criteria. Cancer-free survival rates at 1 and 3 years were 93% and 86% for focal HGD compared with 62% and 44% for diffuse HGD (P < 0.001). On univariate analysis, extent of HGD (relative risk, 5.36; 95% confidence interval, 1.84-15.56), nodularity on endoscopy (relative risk, 3.98; 95% confidence interval, 1.97-8.04), and lack of acid suppression (relative risk, 2.48; 95% confidence interval, 1.16-5.28) were associated with an increased risk of esophageal adenocarcinoma. Diffuse HGD had a 3.7-fold increase in the risk of esophageal cancer compared with focal HGD (P = 0.02) on multivariate analysis.
Conclusions: Patients with focal HGD are less likely to have cancer during the first year after diagnosis or on subsequent follow-up compared with diffuse HGD.