3.1

Introduction

Barrett’s esophagus (BE) occurs in the setting of chronic gastroesophageal reflux disease (GERD). It is characterized by conversion of the normal esophageal squamous epithelium into metaplastic columnar epithelium . Since its original description by Tileston et al. there have been several alterations of this definition. It is now widely accepted that BE is an alteration of the esophageal mucosa that is visible endoscopically and has a corresponding histologic abnormality. The endoscopic aspects about recognizing BE are reviewed in greater detail in Chapter 5 : “Diagnosis of Barrett’s Esophagus.” The discussion here focuses on a brief review of the definition of BE, intestinal metaplasia at gastroesophageal junction (GEJ), and the gross and microscopic aspects of BE and BE-related dysplasia.

3.2

Normal Anatomy and Histology

GEJ, as defined in North America, is the junction between the tubular esophagus and proximal stomach. The location of GEJ is approximated by the most proximal extent of the gastric folds . In Asia, this is determined by locating the distal extent of palisade of longitudinal veins . The squamocolumnar junction (SCJ), also known as the Z-line, is the junction of the squamous mucosa and columnar mucosa. It is normally somewhat irregular in appearance. The location of SCJ and GEJ may not coincide endoscopically. Furthermore, the precise anatomic localization remains difficult in many cases, particularly in the setting of a hiatal hernia . Thus, the precise location of the biopsy is crucial in determining the presence of BE.

Traditionally, the narrow segment of mucus-secreting columnar mucosa distal to the squamous esophageal mucosa, but proximal to acid-secreting gastric oxyntic mucosa has been termed the gastric cardia. In recent years, the existence of the gastric cardia as a native structure has been called into question by some authors who believe that cardiac-type mucosa is always metaplastic, likely in response to gastroesophageal reflux . While metaplastic cardiac-type mucosa undoubtedly is frequently identified in the distal esophagus, evidence from detailed studies of the anatomy and histology of the GEJ, including pediatric autopsy series, supports the notion that the gastric cardia is a native structure . Thus, there is sufficient evidence to support the presence of a small zone of native cardiac mucosa in the most proximal stomach, and in many individuals, metaplastic cardiac-type mucosa of variable length in the distal esophagus.

3.3

Histology of Barrett’s Esophagus

Although Tileston described the presence of reflux-induced esophageal ulcers in an autopsy series, it was not until nearly 50 years later that Dr Norman Barrett confirmed that this was indeed the lower esophagus that was lined by columnar epithelium . However, it was not until later in 1976 that Paull et al. reported that columnar metaplasia of esophagus (or columnar-lined esophagus (CLE)) in fact is a mosaic of three different types of epithelia: (a) fundic type (with oxyntic glands), (b) junctional type (with cardiac-type glands), and (3) specialized type (with goblet cells) . Multiple studies have documented that the risk of dysplasia and adenocarcinoma is significantly increased when the metaplastic epithelium harbors intestinal metaplasia. For this reason, in the United States, we currently diagnose BE based on the 2008 American College of Gastroenterology Practice Parameters Committee Guidelines and the recent Position Statement issued by the American Gastroenterological Association . Per these guidelines, BE is defined as “change in the distal esophageal epithelium of any length that can be recognized as columnar-type mucosa at endoscopy and is confirmed to have intestinal metaplasia by biopsy of the tubular esophagus” . Thus, the definition consists of a two-pronged approach—an endoscopically visible mucosal change/abnormality and a histologic correlate characterized by the identification of goblet cells. This definition, however, is not universally accepted. In fact, the British Society of Gastroenterology does not require the presence of intestinal metaplasia to diagnose BE . Similarly, the Japanese require the documentation of CLE with or without goblet cells . Additional details regarding evolving definition of Barrett’s esophagus can be found in Chapter 5 : “Diagnosis of Barrett’s Esophagus.” An international agreement on a definition of Barrett’s was achieved in 2015 by the BOB CAT (Benign Barrett’s and CAncer Taskforce) consensus group for the first time as “Barrett’s is defined by the presence of columnar mucosa in the esophagus and it should be stated whether intestinal metaplasia is present above the gastroesophageal junction” .

3.3.3

Goblet Cells versus Pseudogoblet Cells

Barrett’s epithelium is typically composed of cells that resemble incomplete intestinal (type II or type III) metaplasia of stomach with chronic gastritis. Complete intestinal metaplasia (type I) is less common. Regardless, subtyping intestinal metaplasia has no clinical implication on the management of BE. Goblet cells are best identified by virtue of their shape and the chemical composition of intracytoplasmic mucin ( Fig. 3.1 ). Because they are rich in acidic mucins (predominantly sialomucins admixed with lesser quantities of sulfated mucins), they acquire a basophilic cytoplasmic blush that is readily recognized on a routine hematoxylin and eosin-stained tissue section . Histochemical stains for acidic mucins, such as Alcian blue at pH 2.5, show intense dark-blue/magenta staining for this combination of sialomucins and sulfated mucins, which contrasts with the predominantly periodic acid–Schiff (PAS)-positive neutral mucins found within the adjacent gastric foveolar-type cells ( Fig. 3.2 ). Routine Alcian blue staining is costly and time-consuming. Since goblet cells can be quite readily recognized on routine hematoxylin and eosin-stained slide, routine use of Alcian blue is not recommended for the diagnosis of BE.

Barrett’s esophagus. The biopsy shows esophageal squamous mucosa that is replaced by glandular mucosa containing goblet cells (arrows).

Periodic acid–Schiff (PAS)/Alcian blue at pH 2.5 demonstrates incomplete intestinal metaplasia. Goblet cells containing acid mucin stain intensely blue with Alcian blue (arrow), while the adjacent columnar cells containing neutral mucin stain with PAS (magenta staining cells within the epithelium).

On occasion, markedly distended foveolar epithelial cells may mimic goblet cells (“pseudogoblet cells”) and cause diagnostic confusion. They are usually distributed in a continuous fashion compared to true goblet cells, which tend to be dispersed sporadically throughout the metaplastic epithelium ( Fig. 3.3 ). Additionally, they stain pale eosinophilic on hematoxylin-eosin stain and contain neutral mucin that does not react with Alcian blue at pH 2.5. Caution must be exercised when interpreting histochemical stains for detecting goblet cells. The columnar cells located between the goblet cells contain small quantities of acidic mucin and may show some Alcian blue positivity (so-called columnar blues). In the absence of goblet cell metaplasia, the identification of these cells does not fulfill the criteria for a definitive diagnosis of BE.

Pseudogoblet cells. These are markedly distended foveolar epithelial cells (arrow), which may mimic goblet cells. They are usually distributed in a continuous fashion compared to true goblet cells, which tend to be dispersed sporadically throughout the metaplastic epithelium.

3.4

Intestinal Metaplasia of the EGJ

The finding of intestinal metaplasia in biopsies obtained from the esophagogastric junction (EGJ) raises two diagnostic possibilities: ultrashort-segment BE or chronic carditis with intestinal metaplasia (CIM). Unfortunately, clinical, endoscopic, or pathologic findings do not allow one to accurately distinguish between these two entities. There is clear evidence that CIM carries a lower risk of neoplastic progression than either short- or long-segment BE In a recent study, Srivastava et al. showed that in a mucosal biopsy from EGJ, the presence of the following features were significantly associated with a diagnosis of BE over CIM: (1) crypt disarray and atrophy, (2) incomplete and diffuse IM, (3) multilayered epithelium, (4) squamous epithelium overlying columnar crypts with IM, (5) hybrid glands, and (6) esophageal glands/ducts .

The expression of CDX2 , a caudal homeobox gene expressed during development, is specific evidence of intestinal differentiation . Several studies have shown that CDX2 is expressed in BE-related intestinal metaplasia . To date, no direct comparisons have been published with regard to potential expression differences between CIM and short-segment BE. Several other markers evaluated for this purpose include Das1, MUC1, MUC2, MUC5AC, MUC6, CK7/CK20 pattern, and CD10. These studies are hampered by differences in endoscopic biopsy protocols and study populations that contribute to apparent discrepancies in their results and lack of reproducibility. Thus, the clinical utility of evaluating intestinal metaplasia of the EGJ using these various biomarkers has not yet been established.

3.5

Barrett’s Esophagus-Related Dysplasia

All patients with BE are at risk of developing esophageal adenocarcinoma . The vast majority of adenocarcinomas arise through a metaplasia-dysplasia-carcinoma sequence. Mapping studies have documented epithelial dysplasia in mucosa adjacent to most adenocarcinomas in resection specimens, supporting a dysplasia-carcinoma sequence . In addition, there are also studies that have reported patients progressing from dysplasia to adenocarcinoma in serial endoscopic biopsies . Histologic evaluation of dysplasia in endoscopic biopsy samples remains the main method of risk assessment in patients with BE. Epithelial dysplasia, particularly high-grade dysplasia, is considered to be one of the most important risk factors for both synchronous and metachronous esophageal adenocarcinoma . Therefore, its identification is an integral part of cancer screening and surveillance programs as well as a trigger point for therapeutic intervention.

Dysplasia is defined as neoplastic change of the epithelium that remains confined within the basement membrane of the gland from which it arises (ie, intraepithelial neoplasia) . Grossly, dysplastic epithelium may demonstrate a spectrum of mucosal changes that ranges from ulcers to flat or elevated/polypoid lesions. This morphologic spectrum forms the basis of the Paris classification used by many gastroenterologists . On occasion, dysplastic mucosa may be indistinguishable from adjacent nondysplastic mucosa.

The most widely accepted histologic grading scheme for Barrett’s-related dysplasia has been adapted from the classification system used for idiopathic inflammatory bowel disease-related dysplasia and is discussed below . In most Western countries, including the United States, Barrett’s-related dysplasia is classified as negative, indefinite, or positive (low grade or high grade). However, pathologists from Asia and Europe prefer the Vienna system of classification. The Vienna system of classifying dysplasia is very similar to the one described earlier, except that the term “noninvasive neoplasia” is used for dysplasia and “suspicious for invasive carcinoma” is used when the cytoarchitectural features are equivocal for tissue invasion. The categories in the Vienna system are: negative for neoplasia, indefinite for neoplasia, noninvasive low-grade neoplasia, noninvasive high-grade neoplasia (includes noninvasive carcinoma in situ and suspicious of invasive carcinoma), and invasive neoplasia (intramucosal and submucosal carcinoma).