Objective: The management of Barrett's high-grade dysplasia (HGD) remains controversial. The aims of this study were to evaluate prospectively the outcome of unifocal HGD (uHGD) in patients with Barrett's esophagus, and to determine demographic and endoscopic features predictive of progression to multifocal HGD (mHGD) and/or adenocarcinoma.
Methods: Consecutive Barrett's patients in whom uHGD was found at initial endoscopy or during surveillance underwent intensification of medical treatment and repeat endoscopy. The study endpoint was progression to mHGD or adenocarcinoma or HGD in conjunction with a dysplasia-associated lesion or mass (DALM). HGD diagnosis was confirmed by a second, blinded pathologist.
Results: A total of 15 Barrett's patients with uHGD met inclusion criteria and have been prospectively followed for a mean +/- SD of 36.8 +/- 23.2 months. All were white and male, with a mean age +/- SD of 61.4 +/- 14.9 yr. Barrett's length varied from 1 to 13 cm (mean, +/- SD, 6.8 +/- 4 cm). Overall, eight (53.3%) uHGD progressed: four of 15 (26.7%) to frank cancer between 17 and 35 months of follow-up, two of 15 (13.3%) to mHGD with DALM in conjunction with one or more foci of possible intramucosal cancer after 12-91 months of follow-up, one of 15 (6.7%) to mHGD with a focus of possible intramucosal cancer after 14 months, and one of 15 (6.7%) to mHGD after 29 months. Seven of 15 (46.7%) uHGD have regressed, five to no dysplasia and two to LGD, over the course of follow-up ranging from 24 to 73 months (mean +/- SD, 43.3 +/- 19.9). All three patients with short-segment Barrett's esophagus with uHGD regressed. Fisher's exact test revealed that Barrett's length > or =3 cm and presence of hiatal hernia approached significance (p < 0.08) in predicting uHGD progression to mHGD/DALM/cancer. However, use of the log-rank test to account for differences in length of follow-up show no significance for hiatal hernia or Barrett's length.
Conclusions: Barrett's uHGD has a high risk for progressing to mHGD or cancer. Justification of an observational approach to uHGD should be discouraged. Markers of uHGD progression, as well as regression, are needed.