Consistency of a high-grade dysplasia diagnosis in Barrett's oesophagus: a Dutch nationwide cohort study.

BACKGROUND: Consistency of high-grade dysplasia in Barrett's oesophagus is incompletely known and the clinical course may vary between patients. AIMS: To evaluate the consistency of high-grade dysplasia diagnosis in a Dutch nationwide cohort and to identify predictors for (re-)detecting high-grade dysplasia or oesophageal adenocarcinoma when ≥ 1 follow-up evaluations after an initial high-grade dysplasia diagnosis were scored with a lower histological grade. METHODS: In this retrospective cohort study, all patients diagnosed with high-grade dysplasia in Barrett's oesophagus between 1999 and 2008 in the Netherlands were selected using the nationwide histopathology registry. Multivariate analysis was performed to identify predictors for (re-)detecting high-grade dysplasia or oesophageal adenocarcinoma in patients with ≥ 1 follow-up evaluations scored with a lower grade. RESULTS: In total, 512 high-grade dysplasia patients were included, of whom 53% had ≥ 1 follow-up evaluations scored with a lower grade. The (re-)detection risk was increased when follow-up was performed in a university hospital and when endoscopic/surgical resection was performed and decreased with an increasing number of follow-up evaluations scored with a lower grade. CONCLUSION: High-grade dysplasia diagnosis was inconsistent in more than half of patients. (Endoscopic) resection in an expert centre is recommended to (re-)detect high-grade dysplasia or oesophageal adenocarcinoma when an endoscopic follow-up protocol with biopsies repeatedly shows a lower histological grade.