Is ablation of the esophageal columnar tissue more likely to induce restoration of squamous epithelium than just acid suppression?
M.M. Berenson (Salt Lake City)
Glandular epithelium that lines the distal esophagus is thought to develop as a consequence of metaplastic re-epithelialization of injured or denuded stratified squamous esophageal mucosa. The glandular epithelium may evolve gradually or rapidly . Bremner et al.  found that regeneration of denuded esophageal mucosa in dogs occurred with squamous epithelium unless reflux of gastroduodenal contents was present in which case glandular tissue developed. They suggested the tissue extended from the gastric mucosa. Presumably, noxious components of the refluxate prevented normal squamous tissue regeneration. Gillen et al.  also found that columnar tissue regenerated in place of denuded stratified squamous esophageal tissue of dogs that refluxed gastric contents, but extension of gastric mucosa into the esophagus was precluded by separating the denuded area of the esophagus from the stomach by a strip of squamous tissue. They suggested the glandular epithelium originated from cells that lined the ducts of the esophageal cardiac glands. Using this canine model, it was subsequently shown that glandular tissue could re-epithelialize denuded areas of squamous epithelium in an anacid environment. Only the presence of islands of squamous tissue mixed into the glandular epithelium differentiated dogs with an anacid environment from dogs with free acid reflux . These observations suggest that esophageal columnar epithelium may develop in the absence of acid reflux. The source of the islands of tissue could be the proximal superficial squamous cells that line the ducts of the esophageal cardiac glands , persistent microscopic foci of normally-differentiated squamous epithelium within the glandular tissue, or primordial stem cells.
Metaplastic tissue has the potential to revert to normal . Both medical [6, 7] and surgical [8-10] therapy to reduce gastric acid secretion, and gastroesophageal reflux have been reported to induce regression of Barrett's epithelium. However, the reports have been infrequent [11-13] and controversial . More commonly, macroscopic and microscopic islands of squamous epithelium have been detected in Barrett's epithelium after antireflux surgery  and long-term acid suppression [11, 12]. Squamous tissue has been observed to grow over glandular epithelium .
We hypothesized that squamous epithelium could be restored to areas of columnar mucosa if the established columnar tissue was ablated and the presence of noxious agents was reduced during the time the esophageal epithelium healed . Argon laser was used to ablate glandular tissue in locations that varied from "islands," totally surrounded by squamous tissue to "patches," totally surrounded by glandular tissue. Gastric acid secretion was suppressed with 40 mg omeprazole daily during the period of re-epithelialization. Normal appearing squamous tissue was partially or totally restored in 38 of 40 treatment locations. No change occurred in Barrett's epithelium that was not treated. Destruction of esophageal glandular tissue using thermal devices [17-21] or photodynamic treatments [22, 23] have subsequently been reported to facilitate squamous tissue re-epithelialization. But, this experience is not without exception . In all cases, antireflux surgery and/or medical treatment to reduce acid secretion and gastroesophageal reflux have been applied during the regenerative period. No study in humans has systematically evaluated the role of acid or other components of the gastroesophageal refluxate on the regenerative process. There are three reasons for which it is generally concluded that anacidity favors normal esophageal tissue regeneration: