SURVEILLANCE IN AN EXPERT BARRETT'S CENTRE ASSOCIATES WITH LOW DISEASE-SPECIFIC MORTALITY IN BARRETT'S OESOPHAGUS PATIENTS

Introduction

Specialist guidelines recommend endoscopic surveillance for Barrett’s oesophagus (BO) to reduce mortality related to oesophageal adenocarcinoma (OAC). The standards for surveillance are poorly defined. We aimed to characterise factors associated with progression and disease-specific mortality in a large cohort of patients followed up in a single Barrett’s expert centre.

Methods

We recruited 980 BO patients to a single-centre prospective study between 1997 to 2021. All endoscopists received training in image-enhanced endoscopy. We identified 869 non-progressors [patients with maximum diagnosis of low-grade dysplasia (LGD) at last follow-up], and 111 progressors [high grade dysplasia (HGD) or OAC during the study period]. Exclusion criteria were: HGD or OAC <12 months from index endoscopy; no intestinal metaplasia or an irregular Z-line; <12 months follow-up, and ablation treatment for LGD. Missed OAC diagnoses in patients lost in follow-up or who declined further surveillance were identified via interrogation of ICD codes and upper GI MDT records (intention-to-treat analysis). Primary outcome measure was progression to HGD/OAC. Secondary outcome measures included cancer stage, type of treatment and disease specific mortality.

Results

The mean surveillance time was 5.65 (range 1.05-18.52) and 7.02 (range 1.00-24.35) years for progressors and non-progressors, respectively. The annual progression rate was 1.65%. On multivariate logistic regression (Table 1) presence of active oesophagitis during surveillance, alcohol use, LGD at baseline endoscopy, male gender, maximum BO length, and hiatus hernia length were risk factors for progression. NSAID use was negatively correlated with progression. 49 patients received an OAC diagnosis, of which 39 (79.6%) had stage I, 5 (10.2%) had stage II, 3 (6.1%) had stage III, 2 (4.1%) had stage IV. Overall, only 4 patients (0.4%) had disease-specific mortality.

Conclusions

Surveillance in an expert Barrett’s centre leads to a high rate of early Barrett’s neoplasia detection and a low rate of disease specific mortality. The identification of specific risk factors could lead to improved prevention and risk stratification tools.