Secondary Logo

Journal Logo

The Natural History of Dysplasia in a Large Cohort of Patients with Barrett's Esophagus


Mathew, Anu MD; Miernik, Artur DO; Gerkin, Richard MD; Ramirez, Francisco C. MD, FACG

American Journal of Gastroenterology: September 2006 - Volume 101 - Issue - p S76
Abstracts: ESOPHAGUS

Gastroenterology, Carl T. Hayde VA Medical Center, Phoenix, AZ and Medicine, Banner Health Good Samaritan Hospital, Phoenix, AZ.

Purpose: For implementing a screening and surveillance program in Barrett's esophagus, it is important to understand the natural history of the disease, in particlar when dysplasia is present.

Aim: To describe the natural history of patients with Barrett's esophagus and any type of dysplasia.

Methods: Endoscopic data from all patients with suspected BE were reviewed. Those patients with intestinal metaplasia and who were found to have dysplasia and at least 1 endoscopic follow up 12 months apart were included. Those undergoing endoscopic therapy for Barrett's were excluded. Patients who presented with adenocarcinoma at index endoscopy were excluded unless there was a prior endoscopy with histology negative for dysplasia.

Study period: January 1997 to December 2004. Setting: Endoscopy Unit at a VAMC.

Results: Of 1714 patients with suspected BE on EGD, 940 had confirmed diagnosis (presence of intestinal metaplasia). Of these, 120 met the criteria with dysplasia (12.8%). These patients had a mean follow up of 4 years.

Nineteen patients (16.2%) had no histological follow up. Fifty nine patients (49%) had negative histologies prior to diagnosis of dysplasia.

When dysplasia was found, regression to no dysplasia and absence of recurrence was seen in 57 (47.5%) cases. Persistent dysplasia (without regression to a lower degree or no dysplasia was seen in 10 patients (8.5%). Two consecutive findings of dysplasia was followed to regression without recurrence of dysplasia in 10 patients (8.5%). Fluctuating dysplasia was observed in 12 cases (10.2%).

There were 13 patients (9.3%) who had adenocarcinoma (5) and/or high grade dysplasia (8) during follow up surveillance.

Tobacco, alcohol, age, and BMI were not significant predictors of histology. The only predictor was Barrett's length. [Figure 1]

Figure 1

Figure 1

Conclusions: 1) In this large, single center cohort of patients with Barrett's esophagus and long follow up, dysplasia of any type (including adenocarcinoma) was found in 12.8% patients. 2) Once found, dysplasia regressed without further recurrence on follow up in 48% cases. 3) Adenocarcinoma and/or HGD developed in 9% patients with dysplasia.

© The American College of Gastroenterology 2006. All Rights Reserved.