Update on the Diagnosis and Treatment of Barrett Esophagus and Related Neoplastic Precursor Lesions

Archives of Pathology & Laboratory Medicine, Oct 2008 by Odze, Robert D

Context.-At present, Barrett esophagus is the most common cause of esophageal adenocarcinoma. In the past 20 years, the incidence of esophageal adenocarcinoma in white males has exceeded that of tumors of the colorectum, lung, prostate, and skin.

Objectives.-To (1) provide an evidence-based review of the diagnosis, classification, and histologic differentiation of Barrett esophagus from gastric carditis, (2) provide a summary of the key pathologic features of precursor lesions, such as dysplasia, and (3) evaluate adjunctive markers of dysplasia and predictive markers for the development of cancer. The natural history and risk of cancer in patients with Barrett esophagus is also reviewed.

Data Sources.-For this review, selected published peer reviewed articles were chosen from a search through PubMed between the years 1970 and 2007.

Conclusions.-The current definition of Barrett esophagus is partially flawed because not all cases are endoscopically recognizable, nongoblet epithelium is biologically intestinalized, and determination of the presence or absence of goblet cells is susceptible to sampling error. Differentiation of ultrashort segment Barrett esophagus from chronic gastric carditis can be accomplished, in a minority of cases, by evaluating for the presence or absence of histologic features that are known to be associated with Barrett esophagus. Dysplasia in Barrett esophagus begins in the crypt bases and then extends more superficially to include the upper portions of the crypts and surface epithelium. Lowand high-grade dysplasia are distinguished by the presence of marked cytologic and/or architectural abnormalities in the latter compared with the former. There are few, if any, reliable adjunctive diagnostic techniques that can help differentiate nondysplastic from dysplastic epithelium. However, α-methylacyl coenzyme A racemase staining has been shown to be useful in 2 separate studies. Both low- and high-grade dysplasia are progressive lesions, and in general, the extent of dysplasia, particularly low grade, is a strong risk factor for progression to carcinoma. Of all the biologic and genetic biomarkers studied to date, evaluation of DNA content is the most reliable and specific. The management of patients with dysplasia is variable among institutions and ranges from aggressive surveillance, endoscopic mucosal resection, mucosal ablation, or total esophagectomy.

(Arch Pathol Lab Med. 2008;132:1577-1585)

In the past 20 years, the incidence of esophageal adenocarcinoma in white males has exceeded that of tumors of the colorectum, lung, prostate, and skin (melanoma).1,2 In the United Kingdom and in whites in the United States, adenocarcinoma is more frequent than squamous cell carcinoma of the esophagus.3 The increase in the annual incidence rate in men is between 10% and 20% in Europe and North America.3-5 Most adenocarcinomas of the esophagus develop on a background of Barrett esophagus (BE), which in turn is caused by chronic gastroesophageal reflux disease (GERD).6 Both GERD and BE are relatively common disorders.7,8 Symptomatic GERD affects approximately 20% of the adult US population, and BE is diagnosed in about 5% of patients who have undergone upper endoscopy for GERD.9,10 The absolute risk of BE patients for cancer is roughly 0.5% per year.1,11

In general, adenocarcinoma in BE develops through a GERD-metaplasia-dysplasia-carcinoma sequence.12,13 Due to the recent widespread use of endoscopy, pathologists are increasingly exposed to biopsy specimens from patients with GERD, BE, and associated dysplasia. Thus, our role as pathologists is to diagnose BE and related neoplastic precursor lesions, and to help guide patient management. This review focuses on recent advances regarding the diagnosis, differential diagnosis, pathogenesis, natural history, and risk assessment of malignancy in BE.

DIAGNOSIS AND CLASSIFICATION OF BE

The American College of Gastroenterology defines BE as endoscopically recognizable columnar metaplasia of the esophageal mucosa that is confirmed to have intestinal metaplasia (defined by the presence of goblet cells) in mucosal biopsy specimens.14 By definition, this disease does not include patients who have intestinal metaplasia of the gastric cardia. Traditionally, BE is separated into long-segment, short-segment, or ultrashort-segment types depending on the length (

As mentioned previously, our clinical colleagues have defined BE by the presence of "intestinal metaplasia" in the esophagus. This is based primarily on the assumption that this epithelium is the type at highest risk for neoplastic progression.19,20 However, there are several weaknesses of this definition. First, it is now well recognized that the background nongoblet columnar epithelium, even in patients without goblet cell metaplasia, shows physiologic properties of "intestinal" differentiation, such as expression of the intestinal transcription factor CDX-2, Hepar-1, villin, DAS-1, and MUC-2.21-25 In addition, nongoblet epithelium in BE also shows cells with a mixed gastric and intestinal phenotype.21 Second, cancer may also develop in goblet cell poor, or rarely even nongoblet, esophageal columnar epithelium.26,27 In fact, both anecdotal evidence and preliminary studies have suggested that neoplastic progression in BE is associated with loss of goblet cell differentiation. In 1 recent study published only in abstract form of 151 BE patients, 63 of whom developed cancer, goblet cell density in nondysplastic epithelium from BE patients without dysplasia was significantly higher than in nondysplastic epithelium from patients with dysplasia.28 In fact, in that study, goblet cell density in nondysplastic crypts was not found to be a useful predictor of cancer risk. The relationship between density of goblet cell metaplasia and cancer risk has never been defined specifically regarding whether this is an absolute (all or none) or a continuous (linear) association. Although it seems logical that an increase in the density of goblet cells is related to an increase in the risk of neoplastic progression, this hypothesis has never been tested objectively. Third, the chance of detecting goblet cells in mucosal biopsies from the esophagus in patients suspected of having BE is essentially proportional to the length of BE and is subject to sampling error.29-31 In a study by Oberg et al,31 the prevalence of goblet cell metaplasia was 30.5% in patients with 1 to 2 cm of columnar-lined esophagus compared with nearly 90% in patients with BE segments longer than 6 cm. Not surprisingly, the likelihood of detecting goblet cells increases with both the number of biopsies and also the location of the biopsy. In fact, biopsies obtained from the neo-squamocolumnar junction shows a higher rate of detection of goblet cell metaplasia compared with biopsies obtained from distal areas of BE. In 1 study of 296 endoscopies in 125 BE patients, goblet cells were detected in 68% of endoscopies in which 8 biopsies were obtained compared with only 35% in which of only 4 biopsies were analyzed.30