What is the role of bile reflux in the pathogenesis of Barrett's esophagus, as shown by experimental models?
S.E.A. Attwood (Salford)
The original animal model which first demonstrated the development of Barrett's esophagus (BE) was the dog model of Bremner . In this study Bremner showed that acid reflux injury is needed for the development of Barrett's columnar lined lower esophagus. He did this by creating a defect in the sphincter at the esophago-gastric junction and then gave repeated injections of histamine to stimulate acid production. The dog, unlike man, produces very little gastric acid between meals. This work has been corroborated by Pollara  and by Gillen . In this dog model no bile reflux was generated and BE was successfully grown. The dog has not been a useful model to study bile reflux as, in these departments, when attempts were made to study bile reflux in the dog, the animals did not survive, due to esophageal necrosis or to stricturing and subsequent aspiration pneumoniae.
The only other study of chronic esophageal reflux injury has been in the rat. Rat models of duodenogastric reflux have not consistently demonstrated the development of columnar epithelium in the lower esophagus. The first study of chronic bile reflux in a rat model was performed by Kranendonk  and supported by work of the author  which showed that, in the rat, refluxed gastric content alone causes very little injury to the squamous epithelium of the rat esophagus. In contrast chronic reflux of duodenal juice in the rat produces a marked inflammatory response in the esophagus and a hyperplastic epithelial response, often with papillomatosis. The most important aspect of the duodenal juice in the generation of this injury is the combination of bile and pancreatic enzymes. Bile on its own has little additional tissue injuring capability above that observed in the gastric reflux model. Subsequent long term studies performed by Clark et al.  and supported by the work of Miwa et al.  have shown occasional islands of columnar epithelium in 10% (4 of 39) rats exposed to chronic bile reflux. On histology these models have not shown circumferential replacement of the lower esophagus with a columnar lining but have been mostly islands of intestinal metaplasia. The importance of this finding is difficult to assess but it has some significance as it is the only model of lower esophageal columar epithelium apart from the original dog model of Bremner.
In a rat model of duodenogastric reflux, administered with a carcinogen, an adenocarcinoma was produced in 50% of the animals [6, 8-10]. While these carcinomas grew in the same model that occasionally produced islands of benign columnar epithelium the animal models have not exactly replicated the development of adenocarcinomas in established benign columnar epithelium, in a way that could be analogous to human Barrett's esophagus. These studies do indicate a possible role for bile or duodenal juices in the generation of adenocarcinoma of the esophagus but there is very little experimental evidence to suggest that bile or duodenal juice reflux is related to the development of a benign Barrett's esophagus.
2. Pollara WM, Zilberstein B, Cecconello I, Filho UL, Pinnotti HW. Regeneration of esophageal epithelium in the presence of gastroesophageal reflux. In: DeMeester TR, Skinner DB, eds. Esophageal disorders: patho-physiology and therapy. New York: Raven Press, 1986.
5. Attwood SEA, Smyrk TC, Marcus J, Murphy B, Steele P, DeMeester TR, Mirvish SS, Hinder RA. Effect of duodenal juice on DNA index and cell proliferation in a model of esophageal carcinoma. Br J Surg 1991;78:A754.
6. Clark GW, Smyrk TC, Mirvish SS, Anselmino M, Yamashita Y, Hinder RA, DeMeester TR, Birt DF. The effect of gastroduodenal juice and dietary fat on the development of Barrett's esophagus and esophageal neoplasia : an experimental rat model. Ann Surg Oncol 1994;1:252-261.
7. Miwa K, Segawa M, Takano Y, Matsumoto H, Sahara H, Yagi M, Miyazaki I, Hattori T. Induction of esophageal and forestomach carcinomas in rats by reflux of duodenal contents. Br J Cancer 1994;70:185-189.
9. Pera M, Cardesa A, Bombi JA, Ernst H, Pera C, Mohr U. Influence of esophagojejunostomy on the induction of adenocarcinoma of the distal esophagus in Sprague-Dawley rats by subcutaneous injection of 2,6-dimethylnitrosomorpholine. Cancer Res 1989;49:6803-6808.
10. Pera M, Trastek VF, Carpenter HA, Fernandez PL, Cardesa A, Mohr U, Pairolero PC. Influence of pancreatic and biliary reflux on the development of esophageal carcinoma. Ann Thorac Surg 1993;55:1386-1393.