Paul Gillen

Regeneration of canine oesophageal mucosa was studied under basal conditions and in the presence of gastro-oesophageal reflux. In normal circumstances mucosal defects in the oesophagus regenerate by squamous epithelium. In the presence of gastro-oesophageal reflux of either acid or a combination of acid and bile, regeneration was frequently by columnar epithelium (Barrett's oesophagus). This columnar regeneration was not seen with bile reflux alone. By the use of squamous barriers to proximal migration of columnar epithelium in the stomach, it was demonstrated that columnar re-epithelialization may occur from cells intrinsic to the oesophagus and is not dependent on proximal migration of cardiac columnar epithelium. The cell of origin of this epithelium may be located in oesophageal gland ducts and is likely to be a multipotential stem cell since the regenerated columnar epithelium may contain goblet and parietal cells not normally found in the oesophagus. This epithelium is morphologically distinct on mucin histochemistry from cardiac columnar epithelium. These findings support the concept that Barrett's epithelium is metaplastic.

Fasting and postprandial intragastric bile acid concentrations have been estimated and compared in patients with complications of Barrett's oesophagus, patients with Barrett's oesophagus without complications, patients with oesophagitis and a group of normal subjects who acted as controls. There was no significant difference in fasting intragastric bile acid concentrations between the groups. Postprandial bile acid concentrations were significantly greater in the patients with complications of Barrett's than in the remaining groups at 60, 90 and 120 min. Significant concentrations of bile acids were seen in gastric juice of unaltered pH and may be undetected on intra-oesophageal pH monitoring. Duodenogastric reflux may be implicated in the pathogenesis of complications of Barrett's oesophagus.

BACKGROUND: This study examined the effect of the lower esophageal milieu on the healing of a mucosal defect. METHODS: With a canine model a mucosal defect was created in the lower esophagus, and acid reflux was promoted by a cardioplasty and pentagastrin injection in 10 animals. At 3 months the regenerated mucosa was excised and underwent histologic examination. Six of these animals underwent an antireflux procedure, had their acid secretion suppressed, and were followed up for a further 3 months, at which time they were killed and the entire esophagus underwent histologic examination. RESULTS: When reflux was stimulated, healing of the mucosal defect was by columnar epithelium alone in seven of 10 animals. After reflux control in six animals healing was again by columnar epithelium, but in all six cases islands of squamous epithelium occurred. Histologic examination showed both types of regenerating epithelium to be in continuity with the ducts of the esophageal glands, lined in their proximal two thirds by columnar and in their distal one third by squamous cells. CONCLUSIONS: It is suggested that the type of regenerating epithelium is determined by the depth of injury to mucosa or gland ducts. If both cell types survive, a mixed pattern of regeneration may occur, but columnar repair will usually predominate because of its more rapid turnover. If the squamous cells of the mucosa and ducts are destroyed, however, repair will be by columnar epithelium alone.

Twenty-four patients with a columnar-lined (Barrett's) oesophagus underwent oesophageal manometry and 24 h ambulatory oesophageal pH monitoring. The results were compared with 25 patients with oesophagitis studied in the same fashion. No significant difference in lower oesophageal sphincter pressure was demonstrated between the two groups. The Barrett's patients demonstrated significantly greater acid exposure in the distal oesophagus than oesophagitis patients. Clearance or refluxed acid was poorer in Barrett's patients than oesophagitis patients. Twelve of the Barrett's patients presented with complications of the condition, i.e. ulceration or stricture. No significant difference in acid exposure was demonstrated between Barrett's patients with or without complications. These results suggest that patients with columnar-lined (Barrett's) oesophagus have greater acid exposure than patients with oesophagitis. The development of complications of a Barrett's oesophagus may not be dependent on acid reflux alone.