Barrett esophagus is named after a London thoracic surgeon who published a paper in 1950 entitled “Chronic Peptic Ulcer of the Esophagus and ‘Oesophagitis.’” In this article, Norman Barrett described cases of esophageal ulcers surrounded by columnar mucosa found at autopsy. Barrett esophagus is the replacement of the normal squamous epithelium of the distal esophagus with specialized intestinal metaplasia (SIM). It is thought to be caused by chronic gastroesophageal reflux disease (GERD), which leads to esophagitis and subsequent metaplastic change of the esophageal lining. SIM may be protective against further injury by gastric acid; however, this metaplastic epithelium is also associated with an increased risk for esophageal adenocarcinoma. There is considerable ongoing debate regarding various aspects of Barrett esophagus, such as the exact neoplastic risk it confers as well as its management if it has become dysplastic.

A. Epidemiology

The overall prevalence of Barrett esophagus in the adult population of the United States is estimated to be 5.6% (although estimates have varied widely from 0.9% to more than 20% depending on the population studied and the definition of Barrett esophagus). As of the 2014 US Census population estimates, this would translate to over 17 million individuals with the condition in the United States. In patients with GERD, the prevalence of Barrett esophagus is higher, approximately 5–10%. In patients with severe GERD, such as those with erosive esophagitis, the prevalence is approximately 10%; in patients with peptic strictures of the esophagus, the prevalence is almost 30%. Barrett esophagus affects males more than females by a ratio of approximately 3:1. The typical patient is a Caucasian, middle-aged male. Although the prevalence of Barrett esophagus in Hispanics appears to be similar to that in Caucasians, Barrett esophagus is uncommon in blacks and Asians. In certain studies, alcohol and smoking have been found to be risk factors for the presence of Barrett esophagus. The conclusion that Barrett esophagus is an acquired complication of chronic GERD is supported by the fact that Barrett esophagus is exceedingly rare in children with competent lower esophageal sphincters (LES).

B. Pathogenesis

The components of the refluxate in GERD (both gastric acid and bile salts) are important etiologic factors in the development of Barrett esophagus. Esophageal pH monitoring studies have shown that patients with Barrett esophagus have more esophageal acid exposure than healthy controls or even patients with mild heartburn. The greater acid exposure in Barrett esophagus results from longer periods of acid reflux (ie, >5 minutes), rather than from a greater number of reflux episodes. Patients with Barrett esophagus may be predisposed to more severe acid reflux episodes due to mechanical dysfunction of the LES as well as the decreased amplitude of distal esophageal contractions.

Helicobacter pylori infection of the stomach causes chronic inflammation that can result in gastric intestinal metaplasia and cancer. However, there is no clear association between H pylori infection and GERD, and the organism does not infect the esophagus. In fact, H pylori infection of the stomach may actually protect the esophagus from GERD and Barrett esophagus.

The role of bile salts in refluxed gastric acid in the development of Barrett esophagus continues to be debated and investigated. Increased levels of refluxed bile acid concentrations have been found in patients with Barrett esophagus. Acid and bile have been shown to induce a class of homeobox genes (Cdx), which are regulators of cell differentiation, and this has been postulated to result in metaplasia. Some clinicians argue that the increased reflux of bile salts in these patients necessitates a mechanical treatment for their GERD, such as a fundoplication. Finally, the genetic and environmental factors that also play a role in the development of Barrett esophagus are only beginning to be understood.

Barrett esophagus cannot be detected on clinical grounds as there are no specific symptoms that distinguish this condition from GERD without Barrett esophagus. In addition, Barrett esophagus cannot be reliably detected radiologically. The gold standard for the diagnosis of Barrett esophagus is upper endoscopy with biopsy of the distal esophagus.

A. Endoscopic and Biopsy Findings

Barrett esophagus is diagnosed when mucosal biopsies confirm the presence of SIM in a columnar-lined distal esophagus. In normal individuals, the squamocolumnar junction (SCJ) coincides with the gastroesophageal junction (GEJ). The SCJ is the transition from the pale squamous mucosa of the esophagus to the salmon-colored gastric mucosa. Endoscopically, the GEJ is best defined as the top of the gastric folds. If there is proximal displacement of the SCJ above the top of the gastric folds into the tubular esophagus, a columnar-lined esophagus is present and biopsies should be taken (Plate 9).

Early studies defined the diagnosis of Barrett esophagus as the presence of 3 cm of columnar-lined distal esophagus with SIM (Plate 10). Subsequent work deemed the presence of less than 3 cm of SIM as “short-segment” Barrett esophagus. Finally, a study in which investigators biopsied the normal-appearing SCJ of consecutive patients undergoing upper endoscopy found a SIM prevalence of 18%. The significance of intestinal metaplasia at the SCJ is debatable, and at present, it is not recommended that a normal-appearing SCJ be biopsied.

It is important to note that intestinal metaplasia of the cardia cannot be distinguished from intestinal metaplasia of the esophagus. Therefore, the proximal displacement of the SCJ into the tubular esophagus must be documented endoscopically in order to diagnose a patient with Barrett esophagus.

The clinical significance of Barrett esophagus length is important as studies have confirmed that longer segments have higher risks of progression to adenocarcinoma. To maintain consistency across studies of Barrett esophagus, the Prague criteria were established to determine the extent of disease. By these criteria, the extent of circumferential (C) and maximal (M) visualized Barrett esophagus is measured, and standard definitions of endoscopic landmarks are used. These criteria have good interobserver reliability for Barrett esophagus ≥ 1 cm in length. It remains to be determined whether these criteria will become adopted broadly.

Biopsy specimens of Barrett esophagus are classified as negative for dysplasia (or non-dysplastic), indefinite for dysplasia (or indefinite-grade dysplasia), low-grade dysplasia, or high-grade dysplasia. The diagnosis of dysplasia is typically reserved for lesions in which nuclear atypia is present on the mucosal surface.

B. Screening

Screening refers to performing a test on a large asymptomatic sample of the population to identify a disease. The role of screening upper endoscopy for Barrett esophagus is unclear, and the major gastroenterologic societies are not unanimous in their recommendations. Patients with chronic GERD are at increased risk for Barrett esophagus, and the prevalence of Barrett esophagus is higher in middle-aged white males, suggesting that screening should be reserved for this higher risk subgroup. The American College of Gastroenterology guideline acknowledges the controversy in screening for Barrett esophagus and only states that the highest yield is achieved in screening patients over age 50 with chronic GERD symptoms.

If the only criteria for Barrett esophagus screening were the presence of GERD symptoms, however, the population to be screened would be enormous as it is estimated that 20% of the US adult population experiences weekly acid reflux. Conversely, a meta-analysis demonstrated that GERD symptoms are not associated with the presence of short-segment Barrett esophagus, but these symptoms do confer a fivefold increase in the risk of long-segment Barrett esophagus.

Decision analysis studies that have examined the cost-effectiveness of screening for Barrett esophagus have shown conflicting results. In addition, a significant proportion of patients with esophageal adenocarcinoma report no prior symptoms of reflux. What cannot be debated, however, is that the majority of patients newly diagnosed with esophageal cancer have not had a previous diagnosis of Barrett esophagus, emphasizing a need for more effective screening strategies. As the cost of esophageal visualization decreases with newer techniques such as transnasal and video capsule endoscopy, these issues will need to be readdressed.