Incidence and progression of Barrett’s oesophagus in clinical practice: A cohort analysis for planning of the Barrett’s ESophagus Trial 3 (BEST3) comparing the Cytosponge-TFF3 test with usual care


Session type:


Judith Offman1,Francesca Pesola1,Rebecca Fitzgerald2,Willie Hamilton3,Peter Sasieni1
1Queen Mary, University of London,2MRC Cancer Cell Unit, University of Cambridge,3University of Exeter



Early detection of oesophageal cancer improves outcomes. BEST3, a cluster-randomised controlled trial, will evaluate the Cytosponge-TFF3 test to identify the pre-cancerous lesion Barrett’s oesophagus (BE) among patients with gastro-oesophageal reflux disease (GERD). It is currently not clear however, how to best identify patients who would not only be eligible for BEST3 but are also most at risk from BE and would therefore benefit the most from the Cytosponge-TFF3 test. The main aim of this study is to aid in the planning of BEST3 by investigating the impact of potential eligibility criteria on the number of eligible patients, and how changes in these eligibility criteria would impact on conversion rates to BE.


A cohort study using data obtained from the Clinical Practice Research Datalink (CPRD), a UK primary care record database, was used to determine the most suitable eligibility criteria by identifying the group of patients with the highest conversion rate to BE. Patients with a prescription of proton pump inhibitors (PPI) or H2 receptor agonists (H2RA) or recorded GERD between 01/01/1987 and 31/12/2010 were included. To estimate conversion rates the number of patients with BE amongst patients with ≥1 endoscopy was calculated.


We found that male compared to female patients (6.2% vs 2.9%), and patients with PPI/H2RA prescriptions compared to without prescriptions (4.7% vs 1.2%) had higher conversion rates to BE. The most suitable inclusion criterion for BEST3 were ≥6 month’s prescriptions of PPI / H2RA with a conversion rate of 6.2% to BE. Furthermore, this criterion identified 46% of GERD patients in the cohort. Variations in PPI / H2RA prescription levels by GP practice ranged from 0.1% to 15% of patients with ≥6 month’s prescriptions per year.


Analysis of this primary care cohort allowed us to identify patients who will benefit most from the Cytosponge-TFF3 test.