Patterns of recurrent and persistent intestinal metaplasia after successful radiofrequency ablation of Barrett's esophagus.

OBJECTIVE: Radiofrequency ablation can eradicate Barrett's esophagus successfully in the majority of cases. We sought to determine (1) how often intestinal metaplasia is detected during follow-up endoscopy after successful ablation and (2) patterns of persistent/recurrent intestinal metaplasia. METHODS: Patients ablated successfully during a phase II clinical trial of radiofrequency ablation for Barrett's esophagus were followed using endoscopic surveillance according to a defined protocol. Systematic biopsies were performed in all patients throughout the neosquamous epithelium as well as at the gastroesophageal junction, and patterns of recurrent or persistent intestinal metaplasia were documented. RESULTS: Fifty-three patients were ablated successfully during this single-institution clinical trial. A total of 151 follow-up endoscopies were performed (range, 1-5 endoscopies per patient) and 2492 biopsies were obtained, of which 604 (24%) were from the gastroesophageal junction. The median follow-up period was 18 months (range, 3-50 months). Recurrent/persistent intestinal metaplasia was detected in 14 patients (26%) in 3 distinct patterns: endoscopically invisible intestinal metaplasia underneath the neosquamous epithelium (buried glands) in 3 patients, visible recurrence in the tubular esophagus in 3 patients, and intestinal metaplasia of the gastroesophageal junction (with a squamous-lined tubular esophagus) in 10 patients. Dysplasia or cancer was not detected in any patient during the follow-up period. CONCLUSIONS: Recurrent/persistent intestinal metaplasia after successful radiofrequency ablation of Barrett's esophagus is relatively common. This finding has implications for the continued surveillance of patients who are ablated successfully.

Effect of hiatal hernia size and columnar segment length on the success of radiofrequency ablation for Barrett's esophagus: a single-center, phase II clinical trial.

OBJECTIVE: Hiatal hernia is common in patients with Barrett's esophagus. We sought to evaluate the effect of hiatal hernia size and initial columnar segment length on the success of radiofrequency ablation of Barrett's esophagus. METHODS: A phase II clinical trial was conducted aimed at evaluating the success of radiofrequency ablation in eradicating Barrett's esophagus. Success was defined as complete replacement of the columnar lining with squamous mucosa and lack of intestinal metaplasia using light microscopy. Hiatal hernia size and columnar segment length were measured endoscopically. RESULTS: Sixty-seven patients were accrued to the protocol. In the 55 patients who completed radiofrequency ablation (43 successes, 12 failures), the mean hiatal hernia size was 3.3 cm (range, 0-10 cm), and the mean columnar segment length was 5.4 cm (range, 1-18 cm). The median length of the columnar segment was 3 cm in the successful cases and 8.5 cm in the failed cases (P = .002). Although the median hiatal hernia size was identical in the successful and failed cases (3 cm, P = .38), the median hiatal hernia size was 7 cm (P = .001) in the 6 patients who experienced nonhealing after the initial ablation. Patients who were successfully ablated but had larger hiatal hernias and longer columnar segment lengths required significantly more radiofrequency ablation sessions than those with smaller hernias and shorter segments (P = .003 and P = .007, respectively). CONCLUSIONS: Patients with larger hiatal hernias and longer columnar segments are more likely to experience failure or nonhealing after radiofrequency ablation. These patients also require more radiofrequency ablation treatments to achieve successful eradication of Barrett's esophagus.

Increased diagnostic yield of small bowel tumors with capsule endoscopy.

BACKGROUND: It is believed that cancers of the small intestine represent <2% of all malignant tumors of the gastrointestinal tract, although the accuracy of this estimate is unknown, because the current methodologies for examining the small bowel have proved inadequate. Capsule endoscopy allows a more detailed inspection of the small intestine and may improve the ability to diagnose small bowel tumors. The objective of this study was to evaluate the effectiveness of capsule endoscopy in diagnosing small bowel tumors and to help establish the true incidence of tumors in obscure gastrointestinal bleeding. METHODS: A retrospective analysis of the charts of 562 patients who underwent capsule endoscopy from August 2001 to November 2003 for a variety of indications was performed. The indication for the procedure was bleeding (alone or in addition to another indication, such as abnormal imaging) in 443 patients. RESULTS: A diagnosis was made by capsule endoscopy in 277 patients (49.3%). Of 562 patients who were included in the study, 50 patients (8.9%) were diagnosed with small bowel tumors. The types of tumor diagnosed by capsule endoscopy included 8 adenocarcinomas (1.4%), 10 carcinoids (1.8%), 4 gastrointestinal stromal tumors (0.7%), 5 lymphomas (0.9%), 3 inflammatory polyps, 1 lymphangioma, 1 lymphangioectasia,1 hemangioma, 1 hamartoma, and 1 tubular adenoma. Of the tumors diagnosed, 48% were malignant. It was observed that 9 of 67 patients (13%) younger than age 50 years who underwent capsule endoscopy for obscure bleeding had small bowel tumors. The pathology results were not available for 10 patients. CONCLUSIONS: Capsule endoscopy diagnosed small bowel tumors in 8.9% of patients who underwent the procedure for a variety of reasons, establishing it as an effective diagnostic modality. This incidence of small bowel tumors suggests an important role for capsule endoscopy in the algorithm for the diagnostic work-up of patients with suspected small bowel lesions. Capsule endoscopy may lead to earlier detection and treatment of small bowel tumors and an improved prognosis for patients with these neoplasms.