This study has two major goals:

1. To determine the effects of bile salts on causing DNA injury and activating signaling pathways that promote growth in cells from the esophagus of patients who have gastroesophageal reflux disease (GERD)
2. To determine whether changes in bile composition induced by treating patients with a bile salt called ursodeoxycholic acid (UDCA) can alter DNA injury, signaling pathway activation and other types of damage in cells from the esophagus of patients who have GERD.

Patients who have been scheduled for elective endoscopic examination at the Dallas VA Medical Center for the evaluation of GERD or Barrett's esophagus will be invited to participate in the study. Patients who provide written, informed consent will have a medical history taken.

Women of child bearing potential will have a pregnancy test. Eligible subjects will be treated with omeprazole 20 mg BID for at least four weeks before the scheduled endoscopic examination. Eight days before the endoscopy, patients will be instructed to discontinue any aspirin and other non-steroidal anti-inflammatory drugs (unless there is a contraindication to discontinuing those medications including a history of coronary artery disease, myocardial infarction, cerebrovascular accident or transient ischemic attacks). The endoscopic examination, which had been scheduled for clinical purposes, will be performed as usual, with biopsy specimens taken as required for clinical purposes.

When the clinical examination has been completed, a perfusion catheter will be passed through the biopsy channel and positioned 5 cm above the squamocolumnar junction in the distal esophagus. The distal esophagus will be perfused with 10cc of a 250 M solution of either deoxycholic acid (DCA) or ursodeoxycholic acid (UDCA) for 5 minutes. Odd-number patients enrolled in each of the two patient groups (GERD patients with and without Barrett's esophagus) will receive DCA, whereas even-number patients will receive UDCA. The catheter position, bile acid concentration and duration of bile acid perfusion are chosen to simulate a typical episode of gastroesophageal reflux.

In all patients, 12 biopsy specimens of the squamous epithelium will be taken using jumbo biopsy forceps at a level 2 cm proximal to the squamocolumnar junction at baseline (6 biopsies will be used to establish the primary cell cultures and six will be used for the molecular analyses); 6 more biopsy specimens will be taken at the same level immediately after bile acid perfusion for molecular analyses.

In the patients with Barrett's esophagus, 12 biopsy specimens of the specialized intestinal metaplasia also will be taken using jumbo biopsy forceps at a level 1 cm distal to the squamocolumnar junction at baseline (6 biopsies will be used to establish the primary cell cultures and six will be used for the molecular analyses); 6 more biopsy specimens will be taken a t the same level immediately after bile acid perfusion for molecular analyses. All endoscopic procedures will be performed by Dr. S.J. Spechler.

All patients will be maintained on omeprazole 20 mg BID for one year, after which the endoscopic examinations will be repeated. The endoscopies will be performed with bile acid perfusions and biopsy sampling exactly as described above, except that patients who received DCA during the first examination will receive UDCA and vice-versa.

After the second endoscopy, patients will be treated with UDCA in a dose of 10 mg/kg for 8 weeks, after which a final endoscopy will be performed.

During this endoscopy, DCA perfusion will be performed as described above. In all patients, 6 biopsy specimens of the squamous epithelium will be taken using jumbo biopsy forceps at a level 2 cm proximal to the squamocolumnar junction at baseline for the molecular analyses; 6 more biopsy specimens will be taken at the same level immediately after bile acid perfusion for molecular analyses.

In the patients with Barrett's esophagus, 6 biopsy specimens of the specialized intestinal metaplasia also will be taken using jumbo biopsy forceps at a level 1 cm distal to the squamocolumnar junction at baseline for the molecular analyses; 6 more biopsy specimens will be taken at the same level immediately after bile acid perfusion for molecular analyses.

Inclusion Criteria:

• Patients who have been scheduled for elective endoscopic examination at the Dallas VA Medical Center for the evaluation of GERD or Barrett's esophagus

Exclusion Criteria:

• Patients unwilling or unable to provide informed consent
• Patients with esophageal carcinomas
• Patients with esophageal varices
• Patients taking warfarin or clopidogrel
• Coagulopathy that precludes safe biopsy of the esophagus
• Comorbidity that precludes safe participation in the study
• Allergy to omeprazole or UDCA
• Pregnancy

3 years