Prior endoscopic surveillance and diagnosis of Barrett-related dysplasia and cancer: evidence from a nationwide Dutch cohort

Background In patients with Barrett oesophagus (BE), surveillance aims to detect dysplasia or oesophageal adenocarcinoma (EAC) early. Whether most asymptomatic patients undergoing endoscopic treatment are identified through surveillance remains uncertain. Objective To assess the contribution of surveillance to the diagnosis of dysplasia and cancer. Design Patients in the Netherlands with curative treatment for BE-related dysplasia or EAC between 2008 and 2018 were identified from the Barrett Expert Center Registry for endoscopic therapy and the Cancer Registry for surgery or chemoradiotherapy. Endoscopy history was based on pathology reports, with at least one endoscopy with non-dysplastic BE 1–5 years before treatment. Primary endpoint was the proportion of patients with dysplasia or EAC with and without (‘de novo patients’) prior endoscopies. Results In total, 8914 patients were included: 90% (95% CI 89% to 91%; 8025/8914) presented de novo, while 10% (95% CI 9% to 11%; 889/8914) had prior endoscopies (mean of 4/IQR 2–7). Overall, 65% had stage ≥II disease, mostly in the de novo group (stage ≥II: 70% de novo vs 11% with prior endoscopy; p<0.001). Two-thirds of endoscopically treated patients (65% (95% CI 62% to 67%); 1283/1989) presented de novo and 97% (95% CI 97% to 98%; 6742/6925) with non-endoscopic therapy. Among those with prior endoscopy, a low rate of missed advanced neoplasia was observed (11%; 96/889). Conclusion In this nationwide cohort, 90% of patients with Barrett-related dysplasia and cancer and 65% undergoing organ-preserving endoscopic treatment had not undergone prior endoscopy. This underscores the need to evaluate whether targeted, evidence-based screening strategies can improve early detection and patient outcomes.