Barrett esophagus is a premalignant condition resulting

Transcription

1 Diagnostic Challenges in the Pathologic Evaluation of Barrett Esophagus Rhonda K. Yantiss, MD N Context. Barrett esophagus represents an unstable epithelium resulting from chronic gastroesophageal reflux disease. Patients with Barrett esophagus routinely undergo endoscopic examination to detect dysplasia and early carcinoma. Although appropriate classification of Barrett esophagus and neoplasia is usually straightforward, persistent esophageal inflammation may induce epithelial changes that mimic, or mask, dysplasia. Recent data also indicate that specific molecular changes occur in nondysplastic Barrett mucosa and herald the development of dysplasia and/or carcinoma. Objective. To describe problematic aspects of biopsy interpretation in tissue samples of the gastroesophageal junction and distal esophagus, including the diagnostic criteria for Barrett esophagus, the importance of the gastric cardia, and pitfalls to the diagnosis of dysplasia. Ancillary studies that have recently emerged as potential Barrett esophagus is a premalignant condition resulting from long-standing gastroesophageal reflux disease. It represents an unstable epithelium that, under the influence of sustained inflammation and injury, is at risk for cumulative molecular alterations, which manifest as progressively severe dysplasia and, ultimately, invasive adenocarcinoma. Patients with Barrett esophagus routinely undergo endoscopic examination to detect dysplasia and early carcinoma, and improved techniques have greatly enhanced gastroenterologists ability to visualize and evaluate neoplastic foci. As a result, Barrett surveillance accounts for a tremendous number of upper endoscopic procedures and mucosal biopsy samples in the United States. Unfortunately, pathologists may face a variety of diagnostic challenges when evaluating biopsies in this setting, particularly since many patients with Barrett esophagus also have persistent gastroesophageal reflux disease, which induces regenerative changes in the glandular epithelium of Accepted for publication February 19, From the Department of Pathology and Laboratory Medicine, Weill Cornell Medical College, New York, New York. Presented in part at the Surgical Pathology of Neoplastic Diseases course, Memorial Sloan-Kettering Cancer Center, New York, New York, May 18 22, The author has no relevant financial interest in the products or companies described in this article. Corresponding author: Rhonda K. Yantiss, MD, Department of Pathology and Laboratory Medicine, Weill Cornell Medical College, 525 East 68th St, New York, NY ( rhy2001@med.cornell. edu). adjuncts to the evaluation of patients with Barrett esophagus will be briefly discussed. Data Sources. A comprehensive review of the relevant literature indexed in PubMed (National Library of Medicine) was performed. Conclusions. Barrett esophagus is currently defined as the presence of intestinal metaplasia in samples obtained from an endoscopically evident abnormality in the distal esophagus. Diagnosis and grading of dysplasia in mucosal biopsies remain the most reliable method to assess risk for neoplastic progression, but its classification may be hindered by superimposed inflammatory changes and suffers from considerable interobserver variability. Therefore, immunohistochemical studies and molecular assessment for TP53, CDKN2A, and DNA content abnormalities have emerged as potential adjuncts to the detection of dysplasia. (Arch Pathol Lab Med. 2010;134: ) the distal esophagus that may mimic, or mask, dysplasia. Thus, the purpose of this review is to describe some problematic aspects of biopsy interpretation that may arise when evaluating samples from patients with Barrett esophagus. Specific topics to be discussed include the diagnostic criteria for Barrett esophagus and intestinal metaplasia, the importance of the gastric cardia, and potential pitfalls in the diagnosis of dysplasia among patients with Barrett esophagus. PROBLEMATIC ASPECTS OF BIOPSY INTERPRETATION IN BARRETT ESOPHAGUS Definition of Barrett Esophagus Barrett esophagus reflects replacement of squamous epithelium of the distal esophagus by glandular mucosa. It is currently defined by the American Gastroenterology Association as (1) an alteration of the distal esophageal epithelium that can be recognized as columnar-type mucosa at endoscopy, in combination with (2) pathologic documentation of intestinal metaplasia (goblet cells). 1 The overall prevalence of Barrett esophagus in the United States is approximately 6%, but it is nearly twice as common among patients with a history of heartburn. Barrett esophagus is classified as long segment (.3 cm), short segment (1 3 cm), and ultrashort segment (,1 cm), depending on the amount of mucosa affected by metaplasia, and both dysplasia and cancer risk are related to increasing length of Barrett esophagus. 2,3 Barrett esophagus often appears as irregular tongues of salmon-pink, velvety mucosa rather than circumferential replacement Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss 1589

2 Figure 1. Barrett esophagus reflects replacement of squamous epithelium by glandular mucosa within the distal esophagus (A). It typically consists of neutral mucin-containing cardiac-type epithelial cells and interspersed goblet cells that contain acid mucins and display cytoplasmic basophilia (B). Goblet cells harbor a large mucin vacuole that compresses the nucleus and adjacent cell borders (C). Samples that display extensive intestinal metaplasia may also contain intestinal-type absorptive cells, Paneth cells, and endocrine cells (D) (hematoxylin-eosin, original magnifications 3100 [A], 3400 [B], and 3600 [C and D]). of squamous epithelium, so some classification schemes use both circumference and length of involvement to determine the affected surface area.4 Biopsy samples from the squamocolumnar junction of patients with Barrett esophagus show neutral mucincontaining columnar (cardiac-type) epithelium (Figure 1, A) with interspersed goblet cells (Figure 1, B). Intestinal metaplasia may be classified as complete, or incomplete, by its appearance and biochemical properties (Table).5 Although incomplete intestinal metaplasia, particularly type III, may portend a greater cancer risk than complete metaplasia, most pathologists do not routinely subclassify intestinal metaplasia in daily practice. Goblet cells are recognizable owing to the presence of mucin-filled vacuoles that distend the cytoplasm, indent the nucleus, and compress the lateral borders of adjacent cells, and their density increases proportionally to the length of Barrett esophagus (Figure 1, C).6 The glands of Barrett esophagus may be irregularly distributed within the mucosa, and display a variable degree of crowding with budding, atrophy, and branching. They often contain cells with enlarged, hyperchromatic nuclei, mucin depletion, and increased mitotic activity, but these features are generally 1590 Arch Pathol Lab Med Vol 134, November 2010 less pronounced on the luminal surface. One may also identify absorptive cells, Paneth cells, and endocrine cells in both the crypt and surface epithelium (Figure 1, D). In addition to cardiac-type epithelium and goblet cells, a variety of other types of glandular mucosa may be observed in gastroesophageal junctional biopsies from patients with chronic reflux disease. Tissue fragments containing fundic (oxyntic) mucosa reflect sampling of the proximal stomach, such as a hiatal hernia, rather than metaplasia within the esophagus and, thus, should not be considered to represent Barrett esophagus. One may also encounter biopsy samples composed of rounded lobules of glands that contain cells with deeply eosinophilic cytoplasm owing to the presence of zymogen granules (Figure 2, A). This finding has been variably termed pancreatic metaplasia or pancreatic heterotopia, but probably represents a congenital variant of normal (pancreatic rest) rather than reflux-related metaplasia (Figure 2, B).7 Some esophageal biopsy samples harbor an epithelium composed of an intimate admixture of basally located squamoid cells and scattered, superficial glandular cells filled with acid mucin (Figure 3, A). This multilayered epithelium may be present either on the luminal surface or Pathology of Barrett Esophagus Yantiss

3 in ducts that drain mucus glands within the mucosa and submucosa (Figure 3, B). Multilayered epithelium is strongly associated with gastroesophageal reflux and concomitant intestinal metaplasia, as well as the development of intestinal metaplasia in subsequent biopsy samples. Glickman et al 8 evaluated multilayered epithelium in samples obtained from 17 patients and found that its glandular cells expressed neutral mucins, sialomucins, and sulfomucins at rates comparable to Barrett esophagus (88%, 100%, and 71% of cases, respectively, versus 100%, 100%, and 76%, respectively). They also noted that glandular cells of multilayered epithelium frequently express CDX2 and MUC2, similar to goblet cells that occur in the esophagus. Therefore, available data suggest that multilayered epithelium is a precursor to intestinal metaplasia of the esophagus, as well as a histologic marker of gastroesophageal reflux disease, although there are no established recommendations regarding the pathologic documentation of multilayered epithelium in biopsy specimens. Both goblet and nongoblet glandular (cardiac-type) epithelial cells within the distal esophagus show immunohistochemical evidence of intestinal differentiation and express CDX2, MUC2, and villin Nongoblet columnar epithelium in the distal esophagus also shows DNA content abnormalities at rates comparable to those of esophageal goblet cells, indicating that esophageal cardiac-type epithelium from patients with Barrett esophagus may already be intestinalized, despite the absence of morphologic goblet cell differentiation. 13 Data from several recent studies 14 indicate that nongoblet glandular epithelium in the distal esophagus may be at risk for neoplastic transformation. Gatenby et al 15 evaluated clinical follow-up for 934 patients, including 322 with intestinal metaplasia and 612 with nongoblet columnar cell metaplasia of the distal esophagus, and found no Classification Scheme of Intestinal Metaplasia Complete Intestinal Metaplasia Incomplete Intestinal Metaplasia Type I Type II Type III Morphology Small intestinal Enterocolonic Colonic Cell types Goblet Absorptive Paneth + +/2 +/2 Intermediate Goblet cell properties Staining qualities Alcian blue High iron diamine Immunophenotype MUC MUC MUC MUC5AC MUC Sulfo-Lewis antigen Intermediate cell properties Mostly sialomucins Mostly sulfomucins Staining qualities Alcian blue NA + + High iron diamine NA 2 + Immunophenotype MUC5AC NA + + MUC6 NA + + Sulfo-Lewis antigen NA 2 + Abbreviations: NA, not applicable; +, positive; 2, negative; +/2, positive or negative. difference with respect to dysplasia and/or cancer incidence between the 2 groups (19.8% versus 15.2%, respectively). In another study, Takubo et al 16 evaluated the histologic features of mucosa adjacent to early adenocarcinomas of the distal esophagus that developed in 141 patients. They found that 57% of patients lacked intestinal metaplasia and that most tumors (70%) developed in cardiac-type or fundic mucosa. These data have led some authorities to propose that the diagnosis of Barrett esophagus need not require the presence of goblet cells. Rather, a growing number of investigators believe that either type of metaplastic glandular mucosa (goblet and nongoblet cardiac-type epithelium) within the distal esophagus should be considered to represent the columnar-lined esophagus, which may be qualified with respect to the presence, or absence, of intestinal metaplasia. 17 Indeed, establishment of a diagnosis of Barrett esophagus in both England and Japan does not require documentation of goblet cells in esophageal biopsies. 18,19 Criteria for the Diagnosis of Intestinal Metaplasia Goblet cells are defined by the presence of acid mucinfilled vacuoles that distend the cytoplasm and impart a blue discoloration to the cytoplasm in hematoxylin-eosin stained sections (Figure 4, A). The acid mucins are deeply basophilic with Alcian blue (ph 2.5) histochemical stains and may show positivity for high iron diamine (Figure 4, B). Their distinction from other types of glandular epithelia may be difficult in some cases. Injured foveolar epithelial cells from patients with active gastroesophageal reflux disease may mimic goblet cells in that they show excessive accumulation of mucin. These cells, which have been termed pseudogoblet cells, are diffusely present in the mucosa and contain mostly neutral, faintly eosinophilic mucin, similar to that of the adjacent uninjured foveolar epithelium Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss 1591

4 Figure 2. Pancreatic heterotopias are commonly encountered in samples obtained from the gastroesophageal junction and appear as rounded lobules of glandular elements (A). The glands contain polarized epithelial cells with basally located nuclei and dense eosinophilic cytoplasm, which reflects the presence of zymogen granules (B) (hematoxylin-eosin, original magnifications 340 [A] and 3400 [B]). Figure 3. Multilayered epithelium may be detected at the luminal surface and/or within deeper glands in gastroesophageal junctional biopsy samples (A). It contains a dual population of basally located squamoid cells admixed with a second population of superficial immature-appearing goblet cells that contain acid mucins (B) (hematoxylin-eosin, original magnifications 3100 [A] and 3600 [B]). (Figure 4, C). However, pseudogoblet cells do contain a small amount of acid mucin, which may show a positive staining reaction with Alcian blue and other histochemical stains.20 The use of special stains to detect goblet cells is, therefore, largely discouraged, since a positive staining reaction does not necessarily imply the presence of goblet cells (Figure 4, D). Acid mucins are also commonly present within epithelium of deep mucosal and submucosal glands and may impart a blue discoloration to the cytoplasm (Figure 4, E). Unlike goblet cells, which are usually interspersed on a background of cardiac-type epithelium, virtually every cell in the gland is affected. Deep glandular cells that contain basophilic mucin also lack cytoplasmic distention and more closely approximate the morphology of cardiac-type epithelial cells (Figure 4, F). Evaluation of Gastroesophageal Junctional Biopsies: The Issue of the Cardia Problems related to the evaluation and diagnosis of Barrett esophagus may be the purview of either gastroenterologists or pathologists. Clinicians are responsible for 1592 Arch Pathol Lab Med Vol 134, November 2010 conveying information regarding the anatomic site of biopsy samples, since mucosal biopsies obtained from within a few millimeters of either side of the gastroesophageal junction may contain cardiac-type glandular mucosa with, or without, goblet cells. Unfortunately, determining the precise site of the biopsy with respect to the gastroesophageal junction may be difficult in some cases. Many patients with chronic gastroesophageal reflux have small hiatal hernias or an irregular Z line, and the definitional criteria for the gastroesophageal junction are variable. Some investigators define the gastroesophageal junction as the proximal tips of the gastric rugal folds, whereas others use a histologic definition, namely the proximal extent of oxyntic mucosa.21 Implicit in the latter definition is the concept that any amount of cardiac-type mucosa near the gastroesophageal junction represents glandular metaplasia of the distal esophagus, rather than a sampling of the proximal stomach, and calls into question the existence of the gastric cardia. The cardia is currently defined as a 1- to 3-mm region that lies between the gastroesophageal junction and Pathology of Barrett Esophagus Yantiss

5 Figure 4. Goblet cells characteristically contain large cytoplasmic vacuoles of acid mucin (A), which show strong Alcian blue positivity (B). Patients with acid-related mucosal injury may have biopsy samples that contain hypertrophic foveolar epithelial cells, which have been termed pseudogoblet cells (C) and typically show some degree of Alcian blue staining (D). Deep mucosal and submucosal glandular epithelial cells may harbor a small amount of acid mucin that imparts blue discoloration of the cytoplasm (E), but, unlike true goblet cells, these mimics (F, top) are diffusely present within the gland and morphologically resemble adjacent cardiac-type epithelial cells (F, bottom) (hematoxylin-eosin, original magnifications 3200 [A and E], 3400 [F], and 3600 [C]; Alcian blue, original magnifications 3200 [B] and 3600 [D]). proximal extent of the gastric folds. Mucosal biopsy samples from this area contain variable numbers of neutral mucin-containing glands associated with a mild amount of chronic inflammation in the lamina propria, Arch Pathol Lab Med Vol 134, November 2010 although some cases may show occasional neutrophils within the surface epithelium. Data from the pediatric population initially suggested that the gastric cardia is a normal anatomic structure present in healthy children, Pathology of Barrett Esophagus Yantiss 1593

6 including those with no history of gastroesophageal reflux disease. 22,23 However, data from adult patients indicate that cardiac-type mucosa near the gastroesophageal junction is likely a metaplastic phenomenon related to chronic inflammation. Chandrasoma et al 24 carefully evaluated the histology of the gastroesophageal junction among 90 autopsy patients and found that the length of cardiac-type mucosa present in this region increased with age. In another study, Oberg et al 25 evaluated gastroesophageal junctional biopsies from 334 adult patients with reflux symptoms, all of whom underwent endoscopic examination and ph probe studies. They found that, when present, cardiac-type mucosa was usually inflamed, and that intestinal metaplasia was associated with abnormal ph studies and inflammation, leading the authors to conclude that the cardia is metaplastic and results from chronic gastroesophageal reflux disease. Others 26 have since shown that any substantial amount of chronic inflammation in glandular mucosa from gastroesophageal junction should prompt concern for the presence of Helicobacter pylori infection, especially when plasma cells are numerous, or when neutrophils are present in both the surface and deep glandular epithelium. Chen et al 27 evaluated 116 mucosal biopsy samples obtained from the vicinity of the gastroesophageal junction and found that progressively severe inflammation was associated with an increased likelihood of H pylori detection. In that study, 57% of patients with severe, chronic active inflammation in cardiac-type mucosa proved to have H pylori infection. However, the likelihood of H pylori detection in biopsies from the proximal stomach decreases with the onset and extent of intestinal metaplasia, since the organism binds via specific bacterial adhesins to cognate host receptors present on foveolar epithelial cells. Goldblum et al 28 assessed mucosal biopsy samples obtained from the gastric side of the gastroesophageal junction and identified 13 patients with intestinal metaplasia of the gastric cardia. Of these, 92% had carditis, 77% had H pylori infection, and 54% had intestinal metaplasia elsewhere in the stomach. Dinçer et al 29 prospectively evaluated 60 patients, 63% of whom were infected with H pylori, and found a negative correlation between the presence of intestinal metaplasia at the gastroesophageal junction and detection of H pylori in the same sample. Therefore, biopsy samples from both the gastroesophageal junction and distal stomach should be obtained during the same procedure to evaluate for H pylori related intestinal metaplasia of the proximal stomach. Intestinal Metaplasia of the Distal Esophagus and Proximal Stomach It is well established that both chronic gastroesophageal reflux disease and H pylori infection may initiate a cascade of events that lead to intestinal metaplasia, dysplasia, and carcinoma, and it is likely that both diseases are responsible for the apparently increasing incidence of gastroesophageal junctional carcinoma in this country. 30 However, the relative contributions of Barrett esophagus and H pylori infection to overall cancer risk are probably unequal: the risk of dysplasia and cancer is highest among patients with long-segment Barrett esophagus, but it is substantially less among patients with either shortsegment Barrett esophagus or intestinal metaplasia limited to the cardia. 31 Thus, there is considerable interest in distinguishing patients with intestinal metaplasia of the proximal stomach from those with Barrett esophagus. Several investigators have used morphologic evaluation and/or immunohistochemical stains to distinguish between esophageal and cardiac intestinal metaplasia. Srivastava et al 32 evaluated mucosal biopsy samples from 20 patients with Barrett esophagus and 20 patients with intestinal metaplasia of the gastric cardia; they found that intestinal metaplasia subjacent to squamous epithelium, intestinal metaplasia confined to the superficial mucosa, and presence of esophageal glands or ducts were only seen in samples from patients with Barrett esophagus (with sensitivity of 57%, 40%, and 30%, respectively). Ormsby et al 33 evaluated the utility of staining patterns for cytokeratins 7 and 20 in distinguishing Barrett esophagus from intestinal metaplasia of the proximal stomach. They found that superficial cytokeratin 20 staining in combination with both superficial and deep cytokeratin 7 staining of goblet cells showed a high sensitivity (97%) and specificity (100%) for Barrett esophagus, as compared to intestinal metaplasia of the gastric cardia, which displayed superficial and deep cytokeratin 20 staining with absent, or weak, cytokeratin 7 staining. Unfortunately, multiple investigators have since failed to reproduce these results. Rather, most have found the Barrett staining pattern to be consistently present in long- and shortsegment Barrett esophagus, as well as intestinal metaplasia in the vicinity of the gastroesophageal junction, but infrequent in intestinal metaplasia of the distal stomach. 10,34 36 Numerous other antibodies, including mucin stains (MUC2, MUC1, and MUC6), hepatocyte antigen, and DAS-1, a colonic epithelial cell protein, have been evaluated for their utility with respect to distinguishing between esophageal, cardiac, and antral intestinal metaplasia. Hepatocyte antigen staining is universally present in intestinal metaplasia of all 3 sites, whereas MUC2 and DAS-1 are usually expressed in intestinal metaplasia of the distal esophagus and proximal stomach (.90%), but are less frequently expressed in the distal stomach. 35 The markers that show the most promise are MUC1 and MUC6, which seem to distinguish between esophageal and cardiac intestinal metaplasia in most cases, although more studies are needed to validate these results. 37 In summary, the major points to consider when evaluating nondysplastic mucosal biopsy samples from patients with suspected Barrett esophagus are as follows. First, although a small amount of cardiac-type mucosa may be detected near the gastroesophageal junction in asymptomatic patients, including children, most investigators now believe that cardiac-type glands in this region are metaplastic and result from subclinical inflammation occurring in 2 main scenarios. Gastroesophageal reflux disease likely accounts for the development of both cardiac-type mucosa and intestinal metaplasia at, or near, the gastroesophageal junction in most cases, but chronic H pylori infection may produce similar changes in the same region. In fact, the prevalence of intestinal metaplasia at the gastroesophageal junction is more than 10% among patients with upper gastrointestinal tract symptoms, at least half of whom do not have gastroesophageal reflux disease. Therefore, the presence of intestinal metaplasia in biopsies labeled gastroesophageal junction may not necessarily reflect Barrett esophagus, but rather, could represent sampling of the proximal stomach. Currently available immunostains are not particularly 1594 Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss

7 helpful in the distinction between intestinal metaplasia of the esophagus and that of the gastric cardia. One simple way to resolve this issue is to perform simultaneous sampling of the distal stomach and gastroesophageal junction, since patients with H pylori related chronic carditis generally have similar changes in the antrum, whereas patients with chronic gastroesophageal reflux disease frequently do not. DYSPLASIA AND CARCINOMA IN BARRETT ESOPHAGUS Classification of Glandular Dysplasia The criteria used for dysplasia classification in Barrett esophagus have, for the most part, been transferred from prior experience with colonic adenomas and inflammatory bowel disease-related colonic dysplasia. Barrett esophagus related dysplasia is evaluated using a 2-tiered grading system (low and high grade), rather than terms such as mild dysplasia, moderate dysplasia, severe dysplasia, and adenocarcinoma in situ. 38,39 Nondysplastic Barrett esophagus shows surface maturation, such that the cells in the deeper glands may contain enlarged nuclei, but accumulate cytoplasm on the mucosal surface (Figure 5, A). The glands are evenly distributed within the lamina propria and often show a mild degree of crowding (Figure 5, B). Scattered mitotic figures, nuclear enlargement, nuclear membrane irregularity, and hyperchromasia may also be evident, but these atypical features are generally mild (Figure 5, C). Most cases of Barrett esophagus related glandular dysplasia display an overall increase in the density of glands within the mucosa and lack surface maturation, similar to colorectal adenomas (Figure 5, D). The same population of atypical cells present in the deep glands is detectable on the luminal surface and is frequently sharply demarcated from nondysplastic epithelium, whereas reactive or inflammatory-type changes tend to be more diffusely present (Figure 5, E). Low-grade dysplasia shows minimal gland architectural abnormalities and mild to moderate cytologic atypia. The neoplastic epithelial cells are crowded and contain elongated nuclei that display hyperchromasia and membrane irregularities. Mitotic figures are readily identified, particularly in the deep glands, but the cells retain their polarity, and nuclei maintain their close proximity to the basement membrane (Figure 5, F). High-grade dysplasia generally displays both gland architectural abnormalities and severe cytologic atypia. Aberrant architectural features include glandular crowding, branching, villiform or cribriform growth, and cystically dilated crypts (Figure 5, G). High-grade dysplastic epithelial cells display complete loss of cell polarity, such that the nuclei lose their relationship to the basement membrane. The nuclei are typically large and ovoid shaped with notched, or irregular, contours and conspicuous nucleoli (Figure 5, H). Both typical and atypical mitotic figures are readily detected at all levels within the glands, as well as on the luminal surface (Figure 5, I). Some cases of high-grade dysplasia display a micropapillary growth pattern and/or contain glands that show considerable variation in size and shape and crowding. In these cases of nonadenomatous dysplasia, the neoplastic cells may assume a cuboidal, or polygonal, shape and contain faintly eosinophilic cytoplasm, vesicular nuclei, and prominent nucleoli, similar to foveolar dysplasia of the stomach. 40 Emerging data also suggest that dysplasia may be detected in the deep glands in the absence of apparent surface maturation. Lomo et al 41 identified 15 patients with Barrett esophagus who were found to have cytologic features of dysplasia limited to the deep glands, in combination with nondysplastic epithelium on the mucosal surface. Eight of these cases showed cytologic features of low-grade dysplasia in the crypt epithelium and 7 showed high-grade cytologic features in this area. Thirteen patients with crypt dysplasia had low-grade dysplasia (2 patients), high-grade dysplasia (8 patients), or intramucosal adenocarcinoma (3 patients) in other biopsies obtained concomitantly, or during prior and/or subsequent procedures. Dysplasia and Cancer Risk in Barrett Esophagus Spontaneous regression of intestinal metaplasia may occur in a minority of patients with Barrett esophagus. Weston et al 42 described complete endoscopic and histologic regression of Barrett esophagus in 7 of 99 patients (7%), which was associated with the absence of a hiatal hernia and decreasing length of Barrett esophagus. Horwhat et al 43 evaluated the natural history of 101 patients with intestinal metaplasia of the esophagus, including 23 patients with long-segment Barrett esophagus, 44 with short-segment Barrett esophagus, and 34 patients with intestinal metaplasia limited to the gastroesophageal junction. They found that all patients with long-segment Barrett esophagus had persistent metaplasia of the distal esophagus despite medical therapy, whereas 30% and 68% of patients with short-segment Barrett esophagus and intestinal metaplasia limited to the gastroesophageal junction, respectively, showed disappearance of goblet cells from the esophageal mucosa. Although these findings suggest that limited intestinal metaplasia of the distal esophagus may occasionally resolve, there are no compelling data indicating that dysplasia spontaneously regresses. Most adenocarcinomas of the esophagus arise in association with Barrett esophagus related dysplasia. Less common scenarios include adenocarcinomas that arise from glandular elements native to the esophagus, such as salivary gland type neoplasms or, rarely, tumors that develop from heterotopic tissues. Secondary involvement of the esophagus may reflect proximal growth of primary gastric cancers into the esophagus, or result from direct extension from the lung or metastatic disease. Older, retrospective studies estimated the prevalence of Barrett esophagus to be about 2% among patients with upper gastrointestinal tract symptoms, and cancer risk to be roughly 10% among patients with Barrett esophagus, although such figures likely reflect cancer prevalence, rather than its incidence among patients followed over time. Recent data indicate that the incidences of dysplasia and cancer development among patients with Barrett esophagus are much lower than previously believed. An intriguing study by Gladman et al 44 evaluated the efficacy of surveillance among 466 patients with Barrett esophagus and found the prevalence of adenocarcinoma to be approximately 6% among these patients. Interestingly, more than 200 patients never returned for surveillance after the diagnosis of Barrett esophagus, and 37 did not have intestinal metaplasia upon review of the initial biopsy sample. All of the remaining 195 patients underwent surveillance for a mean of 5.5 years, during which Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss 1595

8 Figure 5. Nondysplastic Barrett mucosa demonstrates surface maturation (A). Surface epithelial cells contain small, basally located nuclei and do not display mitotic activity, despite the presence of mild cytologic atypia within the crypts (B). Nondysplastic crypts may show glandular crowding, increased nuclear size, and conspicuous nucleoli with occasional mitotic figures (C). In contrast, low-grade dysplasia lacks surface maturation (D). Foci of dysplasia are often sharply demarcated from adjacent nondysplastic epithelium (E) and contain cells with enlarged, hyperchromatic nuclei and scattered mitotic figures (F). High-grade dysplasia frequently displays both architectural and cytologic abnormalities. Neoplastic mucosa may appear villous and contain fused or cribriform glands (G), or harbor micropapillary projections and tufts of highly atypical epithelial cells that show crowding and a complete loss of cell polarity (H). Mitotic figures and necrotic cells are identifiable at all levels within the mucosa, and nuclei show markedly irregular contours, heterogeneous chromatin, and enlarged nucleoli (I) (hematoxylin-eosin, original magnifications 3200 [A, D, and G], 3400 [B, E, and H], and 3600 [C, F, and I]) Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss

9 3% had low- or high-grade dysplasia and 2% had invasive adenocarcinoma. Emerging evidence suggests that low-grade dysplasia may progress to more severe disease in some cases. Skacel et al 45 evaluated biopsies from 25 patients with Barrett esophagus who carried a diagnosis of low-grade dysplasia, which was agreed upon by 10 general surgical pathologists and reviewed by 3 pathologists with an interest in gastrointestinal pathology. When none of the subspecialists agreed with the interpretation of dysplasia, none of the patients had disease that progressed to highgrade dysplasia or carcinoma, whereas increasing concordance among the 3 experts was associated with a higher likelihood of identifying dysplasia or carcinoma in later biopsies. In another study, 46 the authors evaluated the natural history of 356 patients with Barrett esophagus, identified during a 21-year period. Thirty-four patients had low-grade dysplasia and, of these, 27% had disease that progressed to high-grade dysplasia or carcinoma during a mean follow-up period of 8 years. Finally, Srivastava et al 47 evaluated the natural history of 77 patients with Barrett esophagus who had dysplasia, including 31 with low-grade and 46 with high-grade dysplasia. Fifty-seven percent of patients had disease that progressed to esophageal adenocarcinoma, and the most reliable predictor of cancer development was the extent of low-grade dysplasia. The association between high-grade dysplasia and carcinoma is also well established. Up to 45% of biopsies that show high-grade dysplasia are associated with concurrent, or subsequent development of, carcinoma. 48 However, in most cases, this figure probably more closely reflects cancer prevalence than incidence. Glandular dysplasia is often undetectable by conventional endoscopic visualization, thereby necessitating the use of random 4-quadrant mucosal biopsies in surveillance protocols or enhancing techniques (eg, acetic acid guided biopsies, magnification chromoendoscopy with indigo carmine staining, high-resolution magnification with narrow-band imaging) Most examples of high-grade dysplasia that appear as an ulcer, nodule, or plaque likely represent superficial samples of adenocarcinoma. 52 Schnell et al 53 followed up 1099 patients with Barrett esophagus for 20 years and identified 79 patients with high-grade dysplasia. Of these, 45 patients were found to have dysplasia during the surveillance period, and 12 had invasive adenocarcinoma. The authors of that study found the overall incidence of high-grade dysplasia to be less than 5% among patients with Barrett esophagus, and that of cancer development to be approximately 16% among patients with established high-grade dysplasia. Overall, the risk of cancer development among patients with Barrett esophagus is estimated to be less than 2%, and is lower among patients with low-grade dysplasia (8% 12%) than those with high-grade dysplasia (16% 31%). 45,53 55 Problematic Aspects of Biopsy Interpretation in Patients With Suspected Dysplasia Diagnostic challenges to the appropriate classification of mucosal biopsy samples from patients with Barrett esophagus are numerous, yet the pathologic diagnosis rendered in Barrett cases affects the future management of patients. Patients without dysplasia, or low-grade dysplasia, are followed up with endoscopic surveillance and 4- quadrant mucosal biopsies every 1 to 2 cm throughout the length of the Barrett esophagus. Those without dysplasia generally undergo reevaluation in 1 to 3 years, whereas low-grade dysplasia, or biopsies that are interpreted to be indefinite for dysplasia, may prompt more frequent surveillance with extensive mucosal sampling. Highgrade dysplasia is a serious diagnosis that will prompt intensive surveillance and may be treated with a variety of modalities including photodynamic therapy, radiofrequency ablation, laser ablation, or argon plasma coagulation These techniques have largely replaced esophageal resection, which is no longer considered primary therapy for superficial disease in many cases, but, rather, is reserved for patients with multifocal dysplasia and/or endoscopically evident nodules or masses. 59 Given that a diagnosis of dysplasia or adenocarcinoma may prompt increased surveillance and/or excision or resection, these diagnoses should be confirmed by 2 pathologists, at least 1 of whom has a particular interest in gastrointestinal pathology. Inflammation is commonly present in samples from patients with chronic gastroesophageal reflux disease and may induce reactive epithelial cell changes that simulate, or mask, dysplasia. Mucosal biopsy samples from patients with Barrett esophagus may be considered to be indefinite for dysplasia when they display atypical features, but do not meet criteria for a diagnosis of dysplasia. Most problematic cases that fall into this category either (1) show cytologic atypia within glands suggestive of lowgrade dysplasia, but also display surface maturation (Figure 6, A) or (2) display low-grade cytologic abnormalities in the surface epithelium, which is also inflamed (Figure 6, B). For example, mucosal erosions and/or inflammation may result in cytologic abnormalities within the glandular epithelium, including mucin depletion, nuclear enlargement and hyperchromasia, nucleolar prominence, and an increased mitotic activity. Infiltration of the surface epithelium by neutrophils may also induce reactive atypia, simulating the appearance of low-grade dysplasia. As a general rule, one should be cautious when making a diagnosis of low-grade dysplasia when cytologic atypia is limited to the crypt region, or when abundant neutrophilic inflammation is present. However, highgrade cytology and architectural abnormalities cannot be adequately explained by either a location within the crypt region (Figure 6, C) or the presence of neutrophils. In fact, high-grade dysplastic lesions are frequently inflamed or even ulcerated (Figure 6, D). Therefore, one may render a diagnosis of high-grade dysplasia despite the presence of apparent surface maturation or neutrophilic inflammation, albeit one would not be faulted for seeking a second opinion in such a situation. Some histologic features present in mucosal biopsies that contain high-grade dysplasia predict the presence of concurrent adenocarcinoma. Zhu et al 60 evaluated surgical resection specimens from 69 patients with Barrett esophagus who carried a diagnosis of high-grade dysplasia based on biopsy analysis. The authors reevaluated the cases and classified the biopsies as showing high-grade dysplasia, high-grade dysplasia with foci suspicious for carcinoma, or invasive carcinoma. They considered architecturally complex, cytologically atypical glandular proliferations to represent high-grade dysplasia if the epithelium appeared to be confined to the basement membrane, whereas the presence of irregular, dilated glands containing necrotic luminal material (Figure 6, E) Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss 1597

10 Figure 6. A diagnosis of indefinite for dysplasia is typically considered in 2 situations. Nuclear enlargement, nucleolar prominence, and mitotic activity limited to the crypt region may raise concerns for low-grade dysplasia, which may be difficult to diagnose in the face of surface maturation (A). Surface epithelial cell atypia that simulates low-grade dysplasia may occur in combination with neutrophilic inflammation (B). However, severe nuclear atypia, mitotic activity, and cellular necrosis should raise suspicion for high-grade dysplasia despite the presence of either apparent surface maturation (C) or neutrophilic inflammation (D). Cystically dilated glands, necrotic epithelial cell debris (E), and confluent growth of dysplastic epithelium (F) in a background of high-grade dysplasia should prompt concern for invasive adenocarcinoma (hematoxylin-eosin, original magnifications 3400 [A, C, D, and F], 3200 [B], and 3600 [E]). or fused sheets of neoplastic cells (Figure 6, F) were considered to be suggestive of, or outright, invasive carcinoma. When they compared the mucosal biopsy interpretations to subsequent surgical resections, the 1598 Arch Pathol Lab Med Vol 134, November 2010 authors found that there was a high correlation between the biopsy findings and those in the resection specimens. Only 1 patient with a biopsy diagnosis of high-grade dysplasia had invasive carcinoma in the resection specipathology of Barrett Esophagus Yantiss

11 men and that carcinoma was limited to the mucosa. On the other hand, when the biopsy diagnosis was suggestive of, or diagnostic of, invasive adenocarcinoma, carcinoma was found in the resection specimen in 72% and 74% of the cases, respectively. These data suggest that, as pathologists, we can probably do better at identifying patients with high-grade dysplasia on biopsy, and separate them from those with invasive adenocarcinoma. 60 FUTURE DIRECTIONS IN THE MANAGEMENT OF BARRETT ESOPHAGUS Molecular Features Barrett esophagus contains clonally expanded, genetically unstable epithelial cells, even in the absence of overt dysplasia. Prospective studies have shown that loss of heterozygosity (LOH), mutations, and CpG island methylation may be detected in nondysplastic Barrett esophagus and, in combination, may predict development of dysplasia and carcinoma. Some patients with Barrett esophagus harbor low-level chromosomal instability in nondysplastic mucosal biopsy specimens and display sensitivity of peripheral blood leukocytes to mutagenic agents, both of which predict progression to greater degrees of chromosomal instability and adenocarcinoma. 64,65 Yu et al 66 obtained high-fidelity histograms on mucosal samples from 187 patients with Barrett esophagus, including 66 without dysplasia, 22 with low-grade dysplasia, 22 with high-grade dysplasia, and 43 with adenocarcinoma. Samples that lacked dysplasia displayed mixed diploid, and mild to moderately aneuploid cell populations in 5% of cases, which were more numerous in cases of dysplasia (60%) and adenocarcinoma (88%). Overall, baseline abnormalities, including 17p LOH, DNA content tetraploidy and aneuploidy, and 9p LOH have been associated with a 79% 10-year cumulative incidence of esophageal adenocarcinoma among patients with Barrett esophagus. 67 Not surprisingly, Barrett esophagus related dysplasia shows similar molecular features, and frequently displays p16 and TP53 mutations, 17p LOH, and aneuploidy. 68,69 These abnormalities may prove to be important biomarkers in the future, although they are of limited clinical utility at the present time. 67 SUMMARY AND CONCLUSIONS Barrett esophagus is currently defined as the presence of intestinal metaplasia (goblet cells) in biopsy samples obtained from an endoscopically evident mucosal abnormality in the distal esophagus. Although Barrett esophagus is a precursor to esophageal adenocarcinoma, a substantial number of patients harbor intestinal metaplasia limited to the vicinity of the gastroesophageal junction, and it is likely that many of these individuals do not require intensive surveillance. For this reason, pathologists should be aware of the spectrum of mucosal changes that may be encountered in biopsy samples obtained from the gastroesophageal junctional region, and avoid overdiagnosis of Barrett esophagus. In addition, pathologic diagnosis and grading of dysplasia in mucosal biopsies remains the most reliable and widely available method to assess potential risk for neoplastic progression and for identifying patients who may require more intensive surveillance or surgical intervention. Unfortunately, the diagnosis of dysplasia may be problematic owing to the presence of superimposed inflammatory changes in biopsy samples obtained from patients with gastroesophageal reflux disease, and the classification of dysplasia suffers from considerable interobserver variability. Therefore, newer ancillary techniques, including immunohistochemical studies and molecular assessment for TP53, CDKN2A, and DNA content abnormalities, have emerged as potential adjuncts to the detection of dysplasia, although most are not yet widely available for routine evaluation of patients with Barrett esophagus. References 1. Wang KK, Sampliner RE. Updated guidelines 2008 for the diagnosis, surveillance and therapy of Barrett s esophagus. Am J Gastroenterol. 2008;103(3): Weston AP, Sharma P, Mathur S, et al. Risk stratification of Barrett s esophagus: updated prospective multivariate analysis. Am J Gastroenterol. 2004; 99(9): Rex DK, Cummings OW, Shaw M, et al. Screening for Barrett s esophagus in colonoscopy patients with and without heartburn. Gastroenterology. 2003; 125(6): Sharma P, Dent J, Armstrong D, et al. The development and validation of an endoscopic grading system for Barrett s esophagus: the Prague C & M criteria. Gastroenterology. 2006;131(5): Bodger K, Campbell F, Rhodes JM. Detection of sulfated glycoproteins in intestinal metaplasia: a comparison of traditional mucin staining with immunohistochemistry for the sulfo-lewis(a) carbohydrate epitope. J Clin Pathol. 2003; 56(9): Oberg S, Johansson J, Wenner J, et al. Endoscopic surveillance of columnarlined esophagus: frequency of intestinal metaplasia detection and impact of antireflux surgery. Ann Surg. 2001;234(5): Wang HH, Zeroogian JM, Spechler SJ, Goyal RK, Antonioli DA. Prevalence and significance of pancreatic acinar metaplasia at the gastroesophageal junction. Am J Surg Pathol. 1996;20(12): Glickman JN, Spechler SJ, Souza RF, Lunsford T, Lee E, Odze RD. Multilayered epithelium in mucosal biopsy specimens from the gastroesophageal junction region is a histologic marker of gastroesophageal reflux disease. Am J Surg Pathol. 2009;33(6): Chaves P, Cardoso P, de Almeida JC, Pereira AD, Leitao CN, Soares J. Nongoblet cell population of Barrett s esophagus: an immunohistochemical demonstration of intestinal differentiation. Hum Pathol. 1999;30(11): DeMeester SR, Wickramasinghe KS, Lord RV, et al. Cytokeratin and DAS-1 immunostaining reveal similarities among cardiac mucosa, CIM, and Barret s esophagus. Am J Gastroenterol. 2002;97(10): Groisman GM, Amar M, Meir A. Expression of the intestinal marker Cdx2 in the columnar-lined esophagus with and without intestinal (Barrett s) metaplasia. Mod Pathol. 2004;17(10): Hahn HP, Blount PL, Ayub K, et al. Intestinal differentiation in metaplastic, nongoblet columnar epithelium in the esophagus. Am J Surg Pathol. 2009;33(7): Liu W, Hahn H, Odze RD, Goyal RK. Metaplastic esophageal columnar epithelium without goblet cells shows DNA content abnormalities similar to goblet cell-containing epithelium. Am J Gastroenterol. 2009;104(4): Kelty CJ, Gough MD, Van Wyk Q, Stephenson TJ, Ackroyd R. Barrett s oesophagus: intestinal metaplasia is not essential for cancer risk. Scand J Gastroenterol. 2007;42(11): Gatenby PA, Ramus JR, Caygill CP, Shepherd NA, Watson A. Relevance of the detection of intestinal metaplasia in non-dysplastic columnar-lined oesophagus. Scand J Gastroenterol. 2008;43(5): Takubo K, Aida J, Naomoto Y, et al. Cardiac rather than intestinal-type background in endoscopic resection specimens of minute Barrett adenocarcinoma. Hum Pathol. 2009;40(1): Riddell RH, Odze RD. Definition of Barrett s esophagus: time for a rethink is intestinal metaplasia dead? Am J Gastroenterol. 2009;104(10): Ogiya K, Kawano T, Ito E, et al. Lower esophageal palisade vessels and the definition of Barrett s esophagus. Dis Esophagus. 2008;21(7): Playford RJ. New British Society of Gastroenterology (BSG) guidelines for the diagnosis and management of Barrett s oesophagus. Gut. 2006;55(4): Chen YY, Wang HH, Antonioli DA, et al. Significance of acid-mucinpositive nongoblet columnar cells in the distal esophagus and gastroesophageal junction. Hum Pathol. 1999;30(12): Chandrasoma P, Makarewicz K, Wickramasinghe K, Ma Y, Demeester T. A proposal for a new validated histological definition of the gastroesophageal junction. Hum Pathol. 2006;37(1): Kilgore SP, Ormsby AH, Gramlich TL, et al. The gastric cardia: fact or fiction? Am J Gastroenterol. 2000;95(4): Derdoy JJ, Bergwerk A, Cohen H, Kline M, Monforte HL, Thomas DW. The gastric cardia: to be or not to be? Am J Surg Pathol. 2003;27(4): Chandrasoma PT, Der R, Ma Y, Dalton P, Taira M. Histology of the gastroesophageal junction: an autopsy study. Am J Surg Pathol. 2000;24(3): Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss 1599

12 25. Oberg S, Peters JH, DeMeester TR, et al. Inflammation and specialized intestinal metaplasia of cardiac mucosa is a manifestation of gastroesophageal reflux disease [discussion in Ann Surg. 1997;226(4): ]. Ann Surg. 1997; 226(4): Wieczorek TJ, Wang HH, Antonioli DA, Glickman JN, Odze RD. Pathologic features of reflux and Helicobacter pylori-associated carditis: a comparative study. Am J Surg Pathol. 2003;27(7): Chen YY, Antonioli DA, Spechler SJ, Zeroogian JM, Goyal RK, Wang HH. Gastroesophageal reflux disease versus Helicobacter pylori infection as the cause of gastric carditis. Mod Pathol. 1998;11(10): Goldblum JR, Richter JE, Vaezi M, Falk GW, Rice TW, Peek RM. Helicobacter pylori infection, not gastroesophageal reflux, is the major cause of inflammation and intestinal metaplasia of gastric cardiac mucosa. Am J Gastroenterol. 2002;97(2): Dinçer D, Besisik F, Sahin E, et al. Intestinal metaplasia of the gastric cardia: a study from Turkey. Hepatogastroenterology. 2002;49(46): Jemal A, Siegel R, Ward E, Hao Y, Xu J, Thun MJ. Cancer statistics, CA Cancer J Clin. 2009;59(4): Hirota WK, Loughney TM, Lazas DJ, Maydonovitch CL, Rholl V, Wong RK. Specialized intestinal metaplasia, dysplasia, and cancer of the esophagus and esophagogastric junction: prevalence and clinical data. Gastroenterology. 1999; 116(2): Srivastava A, Odze RD, Lauwers GY, Redston M, Antonioli DA, Glickman JN. Morphologic features are useful in distinguishing Barrett esophagus from carditis with intestinal metaplasia. Am J Surg Pathol. 2007;31(11): Ormsby AH, Goldblum JR, Rice TW, et al. Cytokeratin subsets can reliably distinguish Barrett s esophagus from intestinal metaplasia of the stomach. Hum Pathol. 1999;30(3): El-Zimaity HM, Graham DY. Cytokeratin subsets for distinguishing Barrett s esophagus from intestinal metaplasia in the cardia using endoscopic biopsy specimens. Am J Gastroenterol. 2001;96(5): Glickman JN, Wang H, Das KM, et al. Phenotype of Barrett s esophagus and intestinal metaplasia of the distal esophagus and gastroesophageal junction: an immunohistochemical study of cytokeratins 7 and 20, Das-1 and 45 MI. Am J Surg Pathol. 2001;25(1): Mohammed IA, Streutker CJ, Riddell RH. Utilization of cytokeratins 7 and 20 does not differentiate between Barrett s esophagus and gastric cardiac intestinal metaplasia. Mod Pathol. 2002;15(6): Glickman JN, Shahsafaei A, Odze RD. Mucin core peptide expression can help differentiate Barrett s esophagus from intestinal metaplasia of the stomach. Am J Surg Pathol. 2003;27(10): Rugge M, Correa P, Dixon MF, et al. Gastric dysplasia: the Padova international classification. Am J Surg Pathol. 2000;24(2): Schlemper RJ, Riddell RH, Kato Y, et al. The Vienna classification of gastrointestinal epithelial neoplasia. Gut. 2000;47(2): Rucker-Schmidt RL, Sanchez CA, Blount PL, et al. Nonadenomatous dysplasia in Barrett esophagus: a clinical, pathologic, and DNA content flow cytometric study. Am J Surg Pathol. 2009;33(6): Lomo LC, Blount PL, Sanchez CA, et al. Crypt dysplasia with surface maturation: a clinical, pathologic, and molecular study of a Barrett s esophagus cohort. Am J Surg Pathol. 2006;30(4): Weston AP, Badr AS, Hassanein RS. Prospective multivariate analysis of factors predictive of complete regression of Barrett s esophagus. Am J Gastroenterol. 1999;94(12): Horwhat JD, Baroni D, Maydonovitch C, et al. Normalization of intestinal metaplasia in the esophagus and esophagogastric junction: incidence and clinical data. Am J Gastroenterol. 2007;102(3): Gladman L, Chapman W, Iqbal TH, Gearty JC, Cooper BT. Barrett s oesophagus: an audit of surveillance over a 17-year period. Eur J Gastroenterol Hepatol. 2006;18(3): Skacel M, Petras RE, Gramlich TL, Sigel JE, Richter JE, Goldblum JR. The diagnosis of low-grade dysplasia in Barrett s esophagus and its implications for disease progression. Am J Gastroenterol. 2000;95(12): Lim CH, Treanor D, Dixon MF, Axon AT. Low-grade dysplasia in Barrett s esophagus has a high risk of progression. Endoscopy. 2007;39(7): Srivastava A, Hornick JL, Li X, et al. Extent of low-grade dysplasia is a risk factor for the development of esophageal adenocarcinoma in Barrett s esophagus [quiz in Am J Gastroenterol. 2007;102(3):694]. Am J Gastroenterol. 2007;102(3): Tharavej C, Hagen JA, Peters JH, et al. Predictive factors of coexisting cancer in Barrett s high-grade dysplasia. Surg Endosc. 2006;20(3): Hoffman A, Kiesslich R, Bender A, et al. Acetic acid-guided biopsies after magnifying endoscopy compared with random biopsies in the detection of Barrett s esophagus: a prospective randomized trial with crossover design. Gastrointest Endosc. 2006;64(1): Sharma P, Marcon N, Wani S, et al. Non-biopsy detection of intestinal metaplasia and dysplasia in Barrett s esophagus: a prospective multicenter study. Endoscopy. 2006;38(12): Anagnostopoulos GK, Yao K, Kaye P, Hawkey CJ, Ragunath K. Novel endoscopic observation in Barrett s oesophagus using high resolution magnification endoscopy and narrow band imaging. Aliment Pharmacol Ther. 2007; 26(3): Montgomery E, Bronner MP, Greenson JK, et al. Are ulcers a marker for invasive carcinoma in Barrett s esophagus: data from a diagnostic variability study with clinical follow-up. Am J Gastroenterol. 2002;97(1): Schnell TG, Sontag SJ, Chejfec G, et al. Long-term nonsurgical management of Barrett s esophagus with high-grade dysplasia. Gastroenterology. 2001;120(7): Reid BJ, Levine DS, Longton G, Blount PL, Rabinovitch PS. Predictors of progression to cancer in Barrett s esophagus: baseline histology and flow cytometry identify low- and high-risk patient subsets. Am J Gastroenterol. 2000; 95(7): Sampliner RE. Updated guidelines for the diagnosis, surveillance, and therapy of Barrett s esophagus. Am J Gastroenterol. 2002;97(8): Ackroyd R, Tam W, Schoeman M, Devitt PG, Watson DI. Prospective randomized controlled trial of argon plasma coagulation ablation vs. endoscopic surveillance of patients with Barrett s esophagus after antireflux surgery. Gastrointest Endosc. 2004;59(1): Barr H, Shepherd NA, Dix A, Roberts DJ, Tan WC, Krasner N. Eradication of high-grade dysplasia in columnar-lined (Barrett s) oesophagus by photodynamic therapy with endogenously generated protoporphyrin IX. Lancet. 1996; 348(9027): Ell C, May A, Pech O, et al. Curative endoscopic resection of early esophageal adenocarcinomas (Barrett s cancer). Gastrointest Endosc. 2007;65(1): Prasad GA, Wang KK, Buttar NS, et al. Long-term survival following endoscopic and surgical treatment of high-grade dysplasia in Barrett s esophagus. Gastroenterology. 2007;132(4): Zhu W, Appelman HD, Greenson JK, et al. A histologically defined subset of high-grade dysplasia in Barrett mucosa is predictive of associated carcinoma. Am J Clin Pathol. 2009;132(1): Lai LA, Paulson TG, Li X, et al. Increasing genomic instability during premalignant neoplastic progression revealed through high resolution array- CGH. Genes Chromosomes Cancer. 2007;46(6): Maley CC, Galipeau PC, Finley JC, et al. Genetic clonal diversity predicts progression to esophageal adenocarcinoma. Nat Genet. 2006;38(4): Maley CC, Galipeau PC, Li X, et al. The combination of genetic instability and clonal expansion predicts progression to esophageal adenocarcinoma. Cancer Res. 2004;64(20): Chao DL, Maley CC, Wu X, et al. Mutagen sensitivity and neoplastic progression in patients with Barrett s esophagus: a prospective analysis. Cancer Epidemiol Biomarkers Prev. 2006;15(10): Paulson TG, Maley CC, Li X, et al. Chromosomal instability and copy number alterations in Barrett s esophagus and esophageal adenocarcinoma. Clin Cancer Res. 2009;15(10): Yu C, Zhang X, Huang Q, Klein M, Goyal RK. High-fidelity DNA histograms in neoplastic progression in Barrett s esophagus. Lab Invest. 2007; 87(5): Galipeau PC, Li X, Blount PL, et al. NSAIDs modulate CDKN2A, TP53, and DNA content risk for progression to esophageal adenocarcinoma. PLoS Med. 2007;4(2):e Galipeau PC, Prevo LJ, Sanchez CA, Longton GM, Reid BJ. Clonal expansion and loss of heterozygosity at chromosomes 9p and 17p in premalignant esophageal (Barrett s) tissue. J Natl Cancer Inst. 1999;91(24): Reid BJ, Prevo LJ, Galipeau PC, et al. Predictors of progression in Barrett s esophagus II: baseline 17p (p53) loss of heterozygosity identifies a patient subset at increased risk for neoplastic progression. Am J Gastroenterol. 2001;96(10): Arch Pathol Lab Med Vol 134, November 2010 Pathology of Barrett Esophagus Yantiss