Abstract

Background Barrett's oesophagus is a common premalignant lesion caused partly by acid reflux. Although the requisite therapy, proton pump inhibitors (PPIs), have been implicated in the progression of Barrett's oesophagus in animal models, harmful effects of prolonged PPI therapy in Barrett's oesophagus is both inconclusive and controversial. We therefore aimed to test the role of PPI-induced hypergastrinaemia in vitro and see whether any biological parameters were useful surrogates of long-term therapy in man.

Methods We undertook detailed serological and tissue assessment of gastrin and CCK₂ receptors in 90 patients randomised to different doses of PPI therapy during a detailed 2-year follow-up. We also undertook a comprehensive study of cell models to study the consequential biological effects of gastrin on the mucosa.

Results Gastrin and its cognate receptor CCK₂R were expressed highest in the stomach, then less in Barrett's oesophagus and least in squamous oesophagus (SqE) (n=20 paired t-test, p<0.01). Analysis of the change in Barrett's oesophagus segment length change in 70 patients who were randomised to high or low PPI dose showed no difference over 2 years (n=70 t-test, p=0.8). Prolonged PPI use did, however, increase the serum gastrin, (36 pg/ml±57 pg/ml to 103 pg/ml±94 pg/ml (paired t test, p<0.05)). In vitro gastrin also induced changes in OE33(E)_cckr Barrett's oesophagus cells, but not OE21(E)_cckr squamous cells, transfected with CCK₂R; migration was induced by 1 ng/ml of gastrin but proliferation only increased with 100 ng/ml (paired t-test, p<0.01) and both were abolished by antagonists.

Conclusion While the short-term effects of gastrin enhance epithelial restitution in Barrett's oesophagus (but not squamous mucosa) there is no clinical evidence that Barrett's oesophagus length expands over time. This study, which is the largest and longest term randomised controlled trial of gastrin biology in Barrett's oesophagus, is further proof of the clinical safety of PPI therapy.