Document resource
Introduction Debate continues on whether gastric metaplasia (GM) alone warrants a diagnosis of Barrett’s oesophagus surveillance entry as currently recommended by UK clinical guidelines. We compared the risk of neoplastic progression to high grade dysplasia/oesophageal adenocarcinoma for individuals with GM and intestinal metaplasia (IM) subtypes detected at index Barrett’s biopsy.Methods Pathology reports for 14,285 Barrett’s oesophagus patients diagnosed with columnar-lined epithelium of the oesophagus ± IM present between 2011 and 2020 were reviewed and categorised into four groups for metaplasia type (GM only, IM only, Mixed (GM and IM) and unknown. Demographic and clinical characteristic difference were assessed by Chi-square. All patients were followed up to end 31st December 2022. Patients with incident cancer of the oesophagus or gastric cardia (including adenocarcinoma and histological unspecified tumours but excluding squamous cell carcinoma) or with high-grade dysplasia of the oesophagus were identified via linkage with the Northern Ireland Cancer Registry (NICR), and deaths were identified by matching with death records from the General Register Office (GRO). Incidence of cancer or HGD was calculated as events per 100 person-years (% per year). Individuals (n=381) with prior OG cancer, HGD at index or unknown emigration status were excluded from progression analysis.Results At index, 39.2% of all Barrett’s oesophagus diagnoses between 2011-2020 had GM only, 28.0% IM only, 14.4% mixed and 18.5% unknown. Differences in demographic and clinical characteristics were observed between metaplasia types. A higher proportion of individuals in IM group were male (65.7% vs 50.5% in GM only group; p<0.05) and patients in GM only group were younger (50% aged <60 years vs 34.7% in IM only group; p<0.05). A higher proportion of IM only and mixed groups had macroscopic and long segment Barrett’s compared to those in GM only and unknown groups.During a total of 88,239 person years of follow-up, 105 patients were diagnosed with oesophageal or gastric cardia cancer and 44 with HGD. The mean person years of follow-up was comparable across metaplasia groups 6.31 (95%CI 6.26-6.36) years. In the entire cohort, incidence of oesophageal, gastric cardia cancer or HGD combined was 0.17% (95% CI 0.14-0.20) per year. For those with GM only at index combined incidence was 0.04% (95% CI 0.02-0.06) per year, for IM only was 0.35% (95%CI 0.28-0.43) per year, for mixed was 0.26% (0.18-0.37) and unknown 0.11% (95%CI 0.07-0.17) per year.Conclusion In one of the largest follow-up studies to date, neoplastic progression in patients with GM only at index biopsy was much lower than those individuals with IM only and mixed metaplasia type. This adds further population-based evidence to reassure patients and clinicians of the low risk of progression for those with gastric metaplasia only present. These findings highlight both the urgent need to further risk-stratify Barrett’s oesophagus patients to optimize surveillance strategies and for accurate sampling and pathological diagnoses at baseline endoscopy.