SEARCH

SEARCH BY CITATION

Keywords:

  • Barrett's esophagus;
  • diagnosis;
  • endoscopic therapy;
  • esophageal adenocarcinoma;
  • surgery

Abstract

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

Interpretation of exploding knowledge about Barrett's esophagus is impaired by use of several conflicting definitions. Because any histological type of esophageal columnar metaplasia carries risk for esophageal adenocarcinoma, the diagnosis of Barrett's esophagus should no longer require demonstration of intestinal-type metaplasia. Endoscopic recognition and grading of Barrett's esophagus remains a significant source of ambiguity.

Reflux disease is a key factor for development of Barrett's esophagus, but other factors must underlie its development, since it occurs in only a minority of reflux disease patients. Neither antireflux surgery nor proton pump inhibitor (PPI) therapy has major impacts on cancer risk. Within a year, a major trial should indicate whether low-dose aspirin usefully reduces cancer risk.

The best referral centers have transformed the accuracy of screening and surveillance for early curable esophageal adenocarcinoma by use of enhanced and novel endoscopic imaging, visually-guided, rather than blind biopsies and by partnership with expert pathologists. General endoscopists now need to upgrade their skills and equipment so that they can rely mainly on visual targeting of biopsies on mucosal areas of concern in their surveillance practice. General pathologists need to greatly improve their interpretation of biopsies.

Endoscopic therapy now achieves very high rates of cure of high-grade dysplasia and esophageal adenocarcinoma with minimal morbidity and risk. Such results will only be achieved by skilled interventional endoscopists. Esophagectomy should now be mainly restricted to patients whose cancer has extended into and beyond the submucosa.

Weighing risks and benefits in the management of Barrett's esophagus is difficult, as is the process of adequately informing patients about their specific cancer risk.


Introduction—an historical perspective

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

The explosion of knowledge and capabilities in the last 20 years

Between 1989 and 1990 this author led a working party on Barrett's esophagus (BE) that reported at the World Congress of Gastroenterology held in Sydney in 1990 and in this journal in 1991.1 The 20 years following this review have seen a large increase of interest in BE, with an associated burgeoning knowledge base.2–5 The quality of research has also increased greatly, from mainly small, uncontrolled, descriptive studies often subject to publication bias,6 to a predominance of well-structured and powered, multi-center, prospective studies which use comparator groups whenever possible. Results of a PubMed search for publications with the term “Barrett's esophagus”, published up to the end of 2009, shown in Fig. 1, chart the explosion of information on this thorny clinical problem. The most recent general reviews on BE are from Shaheen and Richter,2 Sharma3 and Spechler et al..4 The emphasis of this article is on recent information that is driving change in the clinical management of BE. For reasons that will be explained, this article defines BE as the presence of esophageal columnar metaplasia of any histologic type or extent.

image

Figure 1. Numbers of publications listed by PubMed for the search term “Barrett's esophagus”, from 1950 to 2009, by decade. From 1980 there has been an approximate tripling of publications over each decade.

Download figure to PowerPoint

History and ambiguities of the term “Barrett's esophagus”

In 1903, Norman Barrett was born at home in Adelaide, just 3 km from this author's office.4,7,8 He moved permanently to England with his family when he was about 10. Barrett, a prolific author,7 is of course best known for his single-author 1950 paper,9“Chronic peptic ulcer of the oesophagus and ‘oesophagitis’ ”. Despite a diligent review of the published literature, Barrett, who was relying primarily on gross pathology, favored an incorrect etiologic interpretation, though he did also discuss what is now believed to be the correct pathogenesis of what others christened as “Barrett's Oesophagus” in 1953.10 In 1957 Barrett belatedly accepted the current basic pathogenetic model and in the same paper made clinicians aware of the association between esophageal adenocarcinoma (EA) and esophageal columnar metaplasia, appropriately crediting others for first suggesting this association.11

Barrett's interpretative stumbles have lead some to the view that his achievement was insufficient to merit “naming rights” for this then obscure, but already noted condition. This is a harsh judgment in the light of the investigative methods available to Barrett in the late 1940s.9 Regardless of whether it is deserved that Barrett's name is attached to this clinical entity, it is now so entrenched that we have it forever: also, whether we use the eponym “Barrett's esophagus” fades into insignificance compared with the need to have the same meaning applied to this term throughout the world, so that the disabling ambiguities that have arisen from use of differing histopathologic and endoscopic definitions become a thing of the past.12,13

Rapid change in management of risk for development of incurable adenocarcinoma

In the 1950s, the cases of EA that Barrett observed in association with BE were advanced, presenting mainly with esophageal obstruction.11 The almost universal presence of metastases at this stage caused a dismal prognosis, even if effective palliation was achieved by esophagectomy. The introduction and remarkable technical evolution of flexible endoscopy has progressively enhanced the ability to diagnose EA at a stage when metastases have not occurred, so that cure became possible with esophagectomy or, more recently, with local endoscopic removal or ablation.2–4

Options and capabilities for diagnosing EA and managing its risks have grown especially rapidly in the last decade, but it remains especially difficult for patients to put levels of risk from their BE in perspective and to balance these accurately with the risks of different management options.14,15 Clinicians also have difficulty with making these relatively complex risk-benefit assessments; their difficulties are compounded by the need to tailor risk management to the needs of each BE patient, when the risks and benefits of management options are changing within a time span of two to three years. Whatever the management decision, it must be carefully weighed against the individual patient-specific risk that BE carries for development of EA that progresses to a point that it is a problem, by either causing disability or death. This judgment point differs from just development of EA. Some enthusiasts for intervention are failing to make such balanced assessments and so expose their patients to unwarranted risk by being inappropriately aggressive in their choice of management.15 Because understanding of the risks and benefits of management options in BE needs considerable background information on recent development in the field of BE, this area is addressed in the final parts of this review. The schema in Fig. 2 shows how interventions on the risk for EA and the processes for its assessment are expected to evolve in the next decade. Substantial changes are likely, with beneficial impacts on management of BE.

image

Figure 2. Strategies and developments of methods that are likely to significantly influence the accuracy and cost-effectiveness of and need for surveillance endoscopy in Barrett's esophagus. Components of the schema with question marks have not yet been shown to be efficacious. Absence of a question mark indicates that these items are already achievable, but still require research on appropriateness or the need to be much more widely implemented than at present. The natural history of esophageal adenocarcinoma needs to inform the steps depicted: development of esophageal adenocarcinoma does not mean that it will necessarily cause harm in an elderly patient who has a short life expectancy.

Download figure to PowerPoint

History, definitions and diagnostic criteria for Barrett's esophagus

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

This section explains the major practical difficulties caused by varying definitions of BE,4,12,13 recent insights that signal a way forward and initiatives already taken to achieve a definition that is accepted world-wide.4,12

The use of differing definitions of “Barrett's esophagus” is a problem that must be fixed

The initial informal consensus definition of BE was the partial replacement of normal esophageal squamous mucosa with metaplastic columnar mucosa. The recognition that the metaplasia was a mosaic of several histologic types in most cases of BE, did not change this basic concept. Over the last 20 years or so, particularly in the USA and Germany, many clinical researchers and opinion-makers have unfortunately been misguided in applying a more restrictive definition of BE to only those individuals in whom intestinal-type metaplasia has been found.12 This change was based on two flawed premises—the illogical opinion that risk for EA should be a requirement for use of the term “Barrett's esophagus”, and, a flow-on from the first premise, that cancer risk was confined to intestinal-type metaplasia, a then unproven and now disproven belief. To their credit, British gastroenterologists have consistently rejected this restrictive definition in both research and clinical practice.2,3 The restrictive “only intestinal-type metaplasia” definition of BE dominates the literature, but many publications use the label of BE either for the non-restrictive “any type of metaplasia” definition, or for varying endoscopic criteria unsupported by histopathology. This is a real problem, since these varying definitions select substantially different cohorts. Yet, most original research papers and reviews imply that data on patients with “BE”, however defined, can be generalized without caveats.

Recent authoritative recommendations for a more specific, evidence-based universal definition

Inputs from the Montréal consensus group

The definition of BE was considered by the Global Evidence-Based Consensus Workshop on the Definition and Classification of Reflux Disease (the Montréal workshop). It was controversially proposed at the workshop that the restrictive definition of BE should be abandoned, for reasons outlined below.12 The Montréal workshop eventually reached consensus that the label “Barrett's esophagus” should be used when any type of esophageal columnar metaplasia is confirmed histologically, with the qualifier whether intestinal-type metaplasia has been found. Though the workshop report,12 has been widely cited, up until very recently there has been no detectable movement away from the use of the restrictive definition in the regions where it is favored.

“Negative for intestinal-type metaplasia” patients are at risk for esophageal adenocarcinoma

At least six persuasive considerations now support the abandonment of the restrictive definition of BE. (1) The illogicality of the requirement that risk for EA should be a defining criterion for BE. (2) The pragmatic point that most endoscopists in routine practice do not take enough biopsies to screen adequately for intestinal-type metaplasia, which to be highly sensitive requires 16 biopsies,16,17 so that many patients are being incorrectly assigned to diagnostic limbo as “not BE” (whatever this means), on the basis of a technically inadequate diagnostic process.17 (3) Even if intestinal-type metaplasia were of paramount importance for cancer risk (which it is not), and is truly absent, the dynamic nature of esophageal columnar metaplasia does not mean that it could not develop over time.18 (4) Abnormal DNA has been found recently to be present to similar degrees in esophageal columnar metaplasia of all types, making the malignant potential of “negative for intestinal-type metaplasia” BE biologically plausible.18,19 (5) More conclusive recent pathologic studies have reported that EA occurs in areas of BE devoid of intestinal-type metaplasia,18 vindicating older, less conclusive studies.12 Most convincing is a meticulous histopathologic analysis by a panel of pathologists especially expert in BE who found that of 174 early EAs removed by endoscopic mucosal resection, 64% had developed in areas of esophageal columnar metaplasia negative for the intestinal type.20 Finally (6) the first crucial data on the natural history of what the Montréal workshop defined as BE have come from a large UK study.21 Consistent with the UK “any metaplasia” definition, absence of intestinal-type metaplasia did not influence whether patients had ongoing endoscopic surveillance; during this, EA developed in the 399 patients in whom intestinal-type metaplasia was not found, at a rate that did not differ significantly from the 379 patients in whom intestinal-type metaplasia had been demonstrated.21

There has just been a major shift of authoritative opinion in the USA on the appropriate definition for Barrett's esophagus

Very recently, Spechler et al.4 have reported that the authors of a recently drafted Technical Review on BE commissioned by the American Gastroenterological Association (AGA) considered the data summarized in the previous paragraph and, as a result, concluded that the “intestinal-type metaplasia only” definition should be discarded, in favor of the following: (Barrett's esophagus is) . . . “the condition in which any extent of metaplastic columnar epithelium that predisposes to cancer development replaces the stratified squamous epithelium that normally lines the esophagus”.4 This circuitously worded definition still clings to the illogical concept that malignant potential should be a requirement for any definition of BE, but this is no longer of practical importance if it is accepted that all types of BE mucosa carry a risk for malignant change. The wording of this new definition conveys more than a whiff of political struggle and will bamboozle some readers, especially those whose first language is other than English, but still, this is real progress! Oddly, at the time of finalizing this article, there is no indication when this review will be published in Gastroenterology, the established journal for publication of AGA Technical Reviews.

Problems associated with the word “Barrett” have influenced choice of terminology

The Montréal workshop12 recommended the term “endoscopically suspected esophageal metaplasia (ESEM)”, pending histological confirmation, rather than “suspected BE”. This was driven by concerns of participants from the USA that the word “Barrett” causes major, unjustified loadings to life insurance premiums.12 This is a real issue that needs to be addressed, but this problem must not influence the clarity of clinical terminology around the world; specifically, it is unjustified to coin yet another code term that would, in time, presumably be discovered and acted on with the same prejudice by insurance companies in the USA. This is a field that needs de-, rather than re-coding!

Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

There is now general acceptance that BE is an acquired abnormality. There remain major gaps in our understanding of factors that lead to its development and the factors that trigger progression to dysplasia and EA.

Reflux disease is a key pathogenetic factor for Barrett's esophagus

Reflux disease is proven to be a major pathogenetic factor for development of BE,2–4 even though in some it is symptomatically mild or silent. In the minority of BE patients with metaplastic segments longer that 3 cm, gastroesophageal competence is usually severely impaired. Limited data also indicate that metaplastic segments shorter than 3 cm are associated with reflux disease.

A careful search for BE (defined as at least 3 cm of metaplasia) in 733 unselected post-mortems revealed seven cases of which only two had been diagnosed during life. After adjustment, this was interpreted as showing that only 1 in 21 cases of BE is recognized during life.2 It is possible, but unlikely, that a different combination of pathogenetic factors operated in at least some cases that were identified only post-mortem.22 The more likely explanation is that these individuals had abnormal levels of reflux, but that this defect caused such low levels of reflux-induced symptoms that their reflux disease never came to medical attention. The fact that reflux disease is asymptomatic in a significant minority, whether BE is present or not, is now well established.

Reflux disease is only part of the pathogenesis of Barrett's esophagus

It is relatively unusual for the observed extent of the metaplastic segment to increase over years, or for BE to appear de novo during clinical observation of reflux disease,2–4 even if this is not treated adequately. This suggests that, in most cases, columnar metaplasia develops abruptly. Since only a minority (around 10–15%) of patients with reflux esophagitis has BE,2–4 there must be additional factors that play important roles in determining the apparently sudden development of BE. One plausible trigger factor is major acute esophageal mucosal injury, superimposed on continuing reflux-induced esophageal mucosal damage. This “double trouble” might prevent normal esophageal mucosal healing with squamous mucosa. Though this is an uninvestigated hypothesis, it has strong indirect support from both observations of esophageal mucosal healing after endoscopic mucosal ablation or resection of areas of esophageal metaplasia and animal models; much data from these scenarios show consistently that the presence or absence of damaging reflux determines whether esophageal mucosa that is severely and acutely injured heals with squamous or metaplastic columnar epithelium.2–4

Several categories of transient, but sometimes very severe esophageal mucosal damage could cause “double trouble”, including, for example, acute infective esophagitis, “pill-induced” esophagitis and the severe stasis “drunkard's esophagitis”. These are speculative causes, but development of esophageal columnar metaplasia has been documented in a patient after severe mucositis caused by cancer chemotherapeutic agents23 and in another with caustic esophageal mucosal injury.24 Unfortunately, neither of these patients was adequately investigated for reflux disease.23,24 Interestingly, in the case of lye-induced injury, BE was confined to the mid-esophagus, presumably the area of most severe acute mucosal lye-induced damage.24 In a long-term follow-up study of achalasia patients treated by a relatively aggressive open esophageal/gastric myotomy, coupled with a Dor patch antireflux procedure, BE was found to develop in 12/67 (18%) of patients.25 This could represent “double trouble” of a different type—chronic esophageal stasis with a superimposed therapy-induced defect of gastroesophageal competence, since all but one of the BE patients had abnormal esophageal acid exposure.

Several systemic factors have been identified which may predispose to columnar metaplastic healing of the previously squamous esophageal mucosa.4 Notable among these are retinoic acid status26 and the effects of genetic heterogeneity, including expression of differing types of insulin growth factor,27 and differences between BE patients and controls in CDX2 expression in the squamous mucosa in response to acid and bile salt-induced injury.28 Factors such as these might place particular individuals in a “double trouble” situation at greater risk for mucosal healing with columnar metaplasia.

Pathogenesis of dysplasia and esophageal adenocarcinoma

The key drivers of the development of dysplasia and EA are unknown. Thinking about likely candidates is currently dominated by the assumption that the hostile esophageal luminal environment in BE patients is a key factor in triggering development of dysplasia and EA.4,26 This paradigm is challenged by the lack of evidence, discussed in more detail later, that major alteration of the luminal environment by therapy is associated with any detectable protection against development of EA. This is most convincing in the case of antireflux surgery.29,30 It does not necessarily follow that because a hostile luminal environment is important in causing the development of BE that this also drives development of EA. It seems more likely that the stability of the mucosa in BE is inherently impaired and that extra-luminal factors such as genetic polymorphisms or being iron-replete could play a more important part in driving EA development than the esophageal luminal environment.4

The ongoing BOSS study offers hope for better definition of factors that drive malignant change in BE.31 This is a huge but well-informed fishing trip in which many variables of possible importance to EA development are being measured in an unprecedented number of BE patients, including those involved in the AspECT study17 (see the section on chemoprevention below). Evaluation of samples includes a genome-wide analysis.

Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

Endoscopy remains the only practical option for the routine diagnosis and assessment of esophageal columnar metaplasia. Despite major technical developments of equipment over the last 20 years, the first steps of endoscopic assessment—the recognition of BE and the grading of its extent—remain a major source of inaccuracy.

The Prague Barrett's C&M criteria for diagnosis and grading of extent of Barrett's esophagus

The first systematic study on the endoscopic recognition of BE, which was carried out by the International Working Group for the Classification of Oesophagitis (IWGCO), resulted in the Prague C & M Criteria.32 These were developed in the light of structured evaluations of the interpretation of purpose-recorded and standardized endoscopic video recordings.

Location of the gastroesophageal junction

The Prague Criteria, illustrated in Fig. 3, have already been widely adopted; they highlight the pivotal first step of the endoscopic diagnosis of BE- the correct determination of the position of the anatomical junction between the stomach and esophagus, a judgment that has bedeviled this field since before Norman Barrett was a schoolboy! The best performing landmark, the position of the tops of the gastric mucosal folds during only modest filling of the stomach and esophagus with air was chosen to define the position of the gastroesophageal junction.32

image

Figure 3. Schema which illustrates the Prague Criteria32 and how to implement them. This schema is available as a free download in color from http://www.iwgco.org and is designed to serve as a reminder about the criteria during endoscopy.

Download figure to PowerPoint

Determination of extent of metaplasia

The Prague Criteria provide a researched, standardized method for description of the extent of BE. The C-value describes the length of circumferential metaplasia, whilst the M (for maximum) value reports the uppermost point of any tongue of metaplasia. These values are referenced to the position of the gastroesophageal junction and are given in centimeters.32

Validation studies of the Prague Criteria

The primary validation of the Prague C&M Criteria showed good inter-observer agreement on the position of the gastroesophageal junction and, as a consequence, good agreement on C&M-values greater than 1 cm. Agreement on presence of metaplastic segments less than 1 cm in length was unacceptably poor. Methods for reliable endoscopic recognition of such “ultra short” segments need to be developed. When an endoscopist who is familiar with the Prague Criteria reports a C or M-value of 1 cm or greater, this should signal a high level of confidence that BE is present; however, when the C or M-value is less than 1 cm, there is major uncertainty about its presence.

A second, independent validation study of the Prague Criteria done in several Asian countries has confirmed the data of the IWGCO and shows that endoscopists can implement these criteria successfully in regions with a low prevalence of BE.33

Insights from training sessions on the Prague Criteria

The Prague Criteria provide a focus for training of endoscopists. This is needed, as special referral centers for BE spend significant amounts of time reversing erroneous diagnoses of BE which cause patients unwarranted concerns and expense. The experience of the IWGCO gathered from live training sessions with endoscopists indicates that most misdiagnoses are attributable to misinterpretation of endoscopic landmarks in patients with a small hiatus hernia, due to failure to spend enough time in observing the region of the diaphragmatic hiatus and the upper end of the gastric mucosal folds at a relatively low level of distension. A self-directed IWGCO video-clip training program on this can be accessed on http://www.iwgco.org.

Why accurate endoscopic location of the gastroesophageal junction is so important

Accurate location of the gastroesophageal junction is diagnostically crucial, since mucosa of columnar appearance above this level has to be concluded to be metaplastic. It is not widely realized that there is no form of detailed endoscopic mucosal imaging, biomarker analysis nor histological examination that has been validated to reliably differentiate between metaplastic esophageal mucosa or the mucosa of the extreme upper stomach (whether or not this mucosa is also metaplastic). Correct interpretation of biopsies at and around the gastroesophageal junction currently depends entirely on the accuracy of the endoscopist in locating biopsies.

Visually directed biopsy for endoscopic diagnosis of dysplasia and early esophageal adenocarcinoma—a new primary strategy

The best practice for endoscopic screening and surveillance of BE has changed radically from “blind” to visually directed biopsy. General endoscopists now need to grapple with this change (Fig. 2).

Look, look again and then first target biopsies to mucosal areas of concern

The Seattle protocol, a standard approach over the last 20 years, was developed because the limited resolution of the endoscopes of the 1980s and 1990s made visual recognition of areas of dysplasia or early EA impossible.34 Taking of multiple closely spaced “blind” (unguided by mucosal surface detail) biopsies from each quadrant at 1 cm intervals was then legitimately considered the most effective primary option for histologic screening of the mucosa.

A recent editorial on the diagnosis of dysplasia and EA by Pech is titled “Declaration of Bankruptcy for Four-Quadrant Biopsies in Barrett's Esophagus?”.35 This attention-grabbing statement, cunningly disguised as a proposal by the question mark, was prompted by convincing evidence that biopsy guided by mucosal appearances is now substantially more sensitive for detection of dysplasia and EA than blind biopsies taken according to the Seattle protocol.36–38 This evolution has been driven by the major improvements of the image resolution of endoscopes that have occurred in the last decade.

Is there still a place for the Seattle protocol?

Guidelines still recommend use of the Seattle protocol as the primary approach to assessment of the mucosa in BE.2,3 These require updating to place greater emphasis on visually guided biopsy with a high-resolution endoscopic system. Currently, it is uncertain how much is added by taking blind, in addition to well-targeted biopsies, but given that general endoscopists are currently inadequately skilled and equipped for recognition of mucosal areas of concern, it is probably best that blind biopsies are also taken at least for the present. In centers expert in BE management use of blind biopsy is now the exception rather than the rule. A shift of emphasis to visually targeted biopsies is likely to substantially improve the sensitivity of surveillance done in routine care; most endoscopists only take a limited number of biopsies,16,17 far fewer than the number required by the Seattle protocol,34 so it is especially important that these few biopsies are first directed at areas of concern.

Which endoscopic mucosal imaging technique should be used in routine practice?

Imaging techniques in use at special BE centers include high-resolution white light magnification endoscopic systems, autofluorescence endoscopes, chromoendoscopy, narrow band imaging (NBI), and in a few places, confocal endomicroscopy.35,38 The relative merits of new and evolving mucosal imaging techniques continue to be assessed with well-designed protocols in special BE centers. Some endoscopic systems combine up to three of these imaging modalities. High-resolution endoscopy is not just a matter of using the “right” endoscope-adequate display of mucosal surface detail also requires a high-resolution video monitor.

General endoscopists need guidance on how they can change their practice and equipment in the immediate future to achieve the best possible results with visually guided biopsy. This is a real challenge. Many endoscopists who have high volume BE referral practices in expert BE centers have such well-trained eyes that they miss only a small proportion of areas of high-grade dysplasia or early EA with white light, high-resolution endoscopy alone.35,38 However, not all expert centers achieve such good results.36

In the case of the less-trained eyes of general endoscopists, dysplasia and early EA will probably not be detected with adequate sensitivity with only high-resolution white light endoscopy. Curvers and Bergman refer to the need for a “red flag” imaging modality that directs the endoscopist who is not a super-specialist in BE to mucosal areas of concern.38 Auto-fluorescence endoscopy is probably the most convincing “red flag” technique, but it is currently unclear how important it is to have this for surveillance carried out in routine endoscopic practice. NBI is a less expensive option that may also be useful as a “red flag” imaging method37,38 which, when used with a high resolution endoscope, also assists with accurate visual targeting of biopsies. Accordingly, the most important initiatives for an effective transition to visually guided biopsies in BE in routine practice should be to better train the eye, as discussed below, to upgrade white light endoscopic systems used for surveillance and to use NBI to help flag and examine mucosal areas of concern. If an auto-fluorescence endoscopic system can be included, this is likely to further improve the accuracy of surveillance by general endoscopists.

Training for endoscopic recognition of dysplasia and early esophageal adenocarcinoma

Maximization of the quality of endoscopic surveillance in BE requires more than enhancements of endoscopic equipment. Endoscopist “eye-training” that complements experience from live endoscopy is essential, since general endoscopists have rare exposure to patients with dysplasia and EA in training and routine clinical practice. One practical solution is “own town” access to well-structured high image quality video-endoscopic training materials. These materials must faithfully capture the images from high-resolution endoscopes without any loss of detail, so that the recording emulates what is seen by the endoscopist during the procedure. Because a video recording that captures everything seen during live endoscopy with a high-resolution endoscope requires storage of very large amounts of data each second, this technology has been developed only very recently. Such systems are being used by the IWGCO in its BORN project. IWGCO members at several major specialist BE centers are making video recordings during use of different imaging modalities according to a carefully developed protocol. The protocol includes correlations of the images with histopathologic findings and these edited materials are being built into a structured self-learning program on recognition of high-grade dysplasia and early EA.38 This resource is expected to be available in late 2011 or early 2012.

“Painting” to identify areas of concern, beyond the use of chromoendoscopy

Chromoendoscopy is a relatively clumsy and poorly reproducible technique that is unsuited for use by general endoscopists as a backup mucosal screening technique. “Spray-on” markers for mucosal areas of concern should not however be dismissed as a possible future option, if what is sprayed on “red-flags” dysplasia or EA with high specificity and sensitivity. Molecular imaging is an intriguing possibility for this; antibodies or other types of molecular probes are sprayed onto the mucosa during endoscopy. The latest of several such candidates for use in BE surveillance is a small peptide sequence which has been shown to bind specifically to the surface of dysplastic mucosa ex vivo; this bound peptide can be detected endoscopically by use of a fluorescent tag.39 Rigorous assessments of the clinical utility of molecular probes for BE surveillance should appear in the next few years. A possible major limitation of using a highly specific probe is the considerable diversity of genetic changes seen in EA4.

In-vivo histology with confocal microscopy—intriguing, but likely to be limited in value by interpretative error

The demonstrated ability of endoscopic confocal microscopy to display mucosal histology is an impressive and provocative technical development. The most recent evaluations of its accuracy suggest that variability of diagnoses among observers is a significant issue,40,41 which is no surprise, given the practical issues of interpretation of traditional biopsies discussed below. Another important limitation of this technique is that it may not provide sufficient opportunity for later review of histologic interpretations by another “pathologist” which, as discussed below, is so important to good management of BE. Considerably more research needs to be done before endoscopic confocal microscopy might be proven as a valid, routine diagnostic approach for mucosal assessment in BE.

Per-endoscopic grading of early esophageal adenocarcinoma

Endoscopic classification and staging by surface topographic features is probably fairly insensitive

Pech and colleagues have used a Japanese surface topographic classification (originally designed to aid recognition of early gastric cancers suitable for endoscopic mucosal resection), to reliably subdivide 380 early EAs into five topographic types. Most BE centers have adopted this classification, but early data suggest that it is not sufficiently sensitive for use as a primary method for defining the clinically crucial depth of penetration of early EA.42

Endoscopic ultrasound is insufficiently sensitive for reliable staging of early esophageal adenocarcinoma

High frequency endoscopic ultrasound is a theoretically attractive option for staging early EA,42 but unfortunately this method has an unacceptably low (less than 30%) sensitivity for detection of submucosal penetration by early, mainly surveillance-detected EA, when histopathologic examination of endoscopically resected EA is used as the gold standard.43,44 These data relegate endoscopic ultrasound to a secondary role for staging apparently early EA.

Endoscopic mucosal resection is the gold standard for staging of esophageal adenocarcinoma

Endoscopic resection presents the pathologist with an extensive “surgical” specimen which gives a highly accurate measure of the extent of EA, both into the depth and along the length of the esophagus. The determination whether EA is confined to the mucosa is the crucial variable, since if this is so, there is only a low risk of metastatic spread: by contrast, penetration of EA into the submucosa to any degree not only makes complete local removal by endoscopic therapy difficult to achieve, but is also associated with a much higher risk of lymph node metastases irrespective of the depth of submucosal penetration.45 If the extent of the original mucosal resection specimen is inadequate for reliable staging, a further endoscopic resection can be done to resolve ambiguity.

Issues of histopathological interpretation of biopsies from Barrett's esophagus

Important practicalities relevant to interpretation of reports on biopsies from BE patients are commonly poorly understood in routine clinical practice. The quality of surveillance endoscopy and decision-making on the need for intervention is frequently impaired by this poor understanding (Fig. 2). A terminological soup unfortunately adds to the interpretative challenge. This article uses the relatively simple terminology of low and high-grade dysplasia and EA for describing biopsy findings, an approach also used in the 1990 review.1 Other categories of dysplasia that have been and still may be used are “indefinite” and “moderate”, but these are not usually considered useful by pathologists working in centers with a special interest in BE.46,47 More recently, “low” and “high grade intraepithelial neoplasm” (LGIN and HGIN) have been introduced in the belief that they are technically better terms that are more consistent with terminology used for other mucosae than “high” and “low grade dysplasia”, respectively. LGIN and HGIN are clumsy terms (like the growing use of “at this time”, instead of “now”), but more importantly, they are confusing for at least gastroenterologists. Probably, as a consequence, these terms are used by only a few, most being pathologists. Use of the abbreviations LGIN and HGIN is worse, since they add to the already daunting code that burdens the understanding of BE-related matters.

The diagnosis “low-grade dysplasia” has a variable pathologist-dependent clinical significance and should be reviewed by an expert

The diagnosis of low-grade dysplasia is unfortunately usually very inaccurate when this is made by pathologists who are not highly expert in BE.47 In one US study, 65% of 20 general pathologists misdiagnosed a case of low-grade dysplasia; 25% classified it as normal and the other 40% as either moderate or high-grade dysplasia, in equal proportions.48 A more recent US study found that general pathologists had only poor to fair interobserver agreement on the diagnosis of low-grade dysplasia (Kappa value 0.32).49 In a study from the Netherlands, 85% of low-grade dysplasia cases diagnosed by general pathologists were downgraded to “not dysplasia” on review by pathologists highly expert in BE.50 This experience, consistent with that of Vieth in Germany,47 highlights the important role that centers expert in BE are playing in refining the diagnosis of low-grade dysplasia.

So, given these diagnostic problems, should the clinician ignore low-grade dysplasia? No—because as explained below, this finding, when confirmed by an expert BE pathologist, should change management, because it indicates a substantially higher risk for EA when compared to those whose BE is diagnosed as free of dysplasia.

A biopsy report of high-grade dysplasia or esophageal adenocarcinoma should first lead to an expert review of the pathology

In the last 20 years there has been no change in the recommendation on the first action to be taken on biopsies reported to contain either high-grade dysplasia or EA1–4,46,47—arrange an expert review of the biopsies! The pathologic diagnosis of high-grade dysplasia or EA has excellent inter-observer agreement among pathologists with extensive experience of BE,46,47 but is far from perfect, especially among general pathologists.48,49 Even expert pathologists differ not infrequently as to whether a biopsy shows high-grade dysplasia or EA,46,47 a discordance that has some clinical significance, since high grade dysplasia, by definition, is confined to the mucosa and so can be cured by endoscopic therapy, rather than esophagectomy; EA carries the risk of submucosal penetration. A recent study has made some recommendations for histopathologic criteria in biopsies that appear to help to distinguish between patients who only have high-grade dysplasia and those who have EA elsewhere in the metaplastic mucosa.51,52 A thorough endoscopic screening of the entire metaplastic mucosa with visually targeted biopsies and mucosal resection specimens should remain the mainstay for guiding management of high-grade dysplasia or early EA.

Epidemiology of Barrett's esophagus

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

The number of high quality studies in this area has increased greatly, especially in the last decade. Epidemiologic data on BE are a rich source of hypothesis-generating information,2–5,53 but need to be interpreted and pooled carefully because of the use of differing endoscopic diagnostic criteria and varying definitions of BE.2–4,15 There is currently strong interest about a possible link between obesity and BE.54 This article needs to concentrate on information about the prevalence of BE and the risk for development of EA, since this is key in molding the management of the EA risk in BE.

Studies of the population prevalence of Barrett's esophagus

Endoscopic surveys of the general population are the only way of measuring the true prevalence of BE; all other approaches provide only “guesstimates”. Two studies, the Kalixanda and SILC studies, both of which involved recruiting and endoscoping ∼1000 volunteers representative of the general population,55,56 have overcome the logistic challenges of such necessarily large endoscopic surveys. Data quality was ensured by the use of a very small number of endoscopists in both studies, who were specially trained in recognition of BE prior to study start. The tops of the gastric mucosal folds were used in both studies to define the position of the gastroesophageal junction.

Prevalence of Barrett's esophagus in Sweden—the Kalixanda Study

The Kalixanda study, done in northern Sweden, reported a surprisingly low 1.6% prevalence of BE for a prosperous Caucasian population,55 when judged against other data derived from post mortem studies and patients referred for endoscopy for clinical indications.2–4 The 1.6% prevalence is misleading, as the main analyses of the Kalixanda study used the restrictive “intestinal metaplasia only” definition; the reported prevalence of this BE subgroup is almost certainly significantly under-estimated, because of the limited number of esophageal biopsies taken,55 inadequate for sensitive screening for intestinal-type metaplasia.16 If endoscopically suspected BE is used instead as the diagnostic criterion, the population prevalence of BE found in the Kalixanda study jumps more than six-fold to 10.3%, a proportion that is probably more meaningful, given the limitations and the poor clinical relevance of the “intestinal metaplasia only” definition discussed above.4,12,13 The Kalixanda study report only gives data on the extent of metaplasia for the 1.6% of intestinal metaplasia positive subjects; only 5 of these 11 (31%) had metaplasia at least 2 cm in length,55 with only 26% of subjects having metaplasia extending 3 cm or greater.

New data from China- the SILC Study

The other large population endoscopic survey, part of the SILC study, was done very recently in Shanghai, China.56 Endoscopically suspected BE was present in 1.9% of subjects. Though the Prague Criteria were applied in this study, no data are given on extent in the published report.56 These data are however available to this author and are given in Table 1: the distribution of extent was similar to the Kalixanda study,55 with only 26% of subjects having a maximum extent of metaplasia 3 cm or greater.

Table 1.  Extent of metaplasia in the 19 subjects with endoscopically diagnosed Barrett's esophagus, according to Prague Criteria in the SILC Study
 C value % of subjects (n)M value % of subjects (n)
Less than 1 cm47 (9)0 (0)
1 to less than 2 cm11 (2)32 (6)
2 to less than 3 cm37 (7)42 (8)
3 cm or greater5 (1)26 (5)

Regional differences of prevalence of Barrett's esophagus and large volume endoscopic surveys of selected populations

The SILC study and the re-interpretation of the Kalixanda study are consistent with the generally held view, derived predominantly from clinical experience, that BE is much more prevalent in relatively prosperous countries with predominantly Caucasian populations, compared to non-Caucasian populations which are usually substantially less prosperous. The reported very low prevalence of EA in non-Caucasian populations is consistent with their reported prevalence of BE, with the notable exception of Japan. There is lively interest in how important genetics are in determining this stark difference compared to environmental factors.53 My money is on environment being dominant!

Health–check endoscopic data from huge numbers of subjects

“Health-check” endoscopy is offered to the general population in several Asian countries, notably Japan, China, Korea and Taiwan. Increased interest in BE in these countries has resulted in five evaluations of the prevalence of variably defined BE in health-check endoscopy subjects.57–61 Details of these studies given in Table 2 show that they have evaluated unprecedentedly huge numbers of subjects, but that they also have some significant technical limitations. The Chinese, Korean and Taiwanese studies found, as expected, a low prevalence of endoscopically suspected BE. More than three quarters of cases had an extent less than 1 cm in the two studies that reported on such an extent (Table 2).57,61 Given that the validation studies on the Prague Criteria showed such poor reproducibility of recognition of metaplastic segments less than 1 cm,32,33 the authority of the great majority of the diagnoses of BE made in these studies appears uncertain, especially as, in the largest series, it is presumed that large numbers of endoscopists were involved. Only one report (Table 2) makes mention of provision of training on recognition of BE to participating endoscopists, but no details are given on this training, nor the criteria applied for diagnosis of BE.61

Table 2.  Summary of surveys for presence and extent of Barrett's esophagus in subjects having health-check endoscopy
CountryFirst authorProtocol designCriteria for G-E JunctionNumber subjectsEndoscopic BEIntest Met Pos BEExtent MetaplasiaRemarks
  • Intest Met Pos BE, Intestinal-type metaplasia demonstrated in biopsies of endoscopically suspected Barrett's esophagus.

  • *

    Not all suspected BE subjects had esophageal biopsies. see “Remarks” column for details.

JapanAkiyama57Stored imagesPrague Criteria86943%Not done98% C valueStored still images evaluated by two experienced endoscopists
<3 cm
Yamagishi58Stored imagesNot stated6 50410.8%Not done84% <1 cmStored still images evaluated by two experienced endoscopists
13% 1–3 cm
3% >3 cm
ChinaPeng59Live endoscopyTops gastric mucosal folds2 580Not stated1.0%*81% <3 cm*Estimated: 74% had biopsies Two experienced endoscopists
TaiwanTseng60Live endoscopyTops gastric mucosal folds19 8120.3%0.07%*87% <3 cm*Estimated: 77% had biopsies Single-center study. Experienced endoscopists-number not stated
KoreaPark61Live endoscopyNot stated25 5363.4%1.1%*78% <1 cm*Estimated: 74% had biopsies Multi-center study Endoscopists numbers not given Pre-study training session
38% 1–1.9 cm
4.6% ≥2 cm

The extraordinarily high and discordant prevalence of BE in the Japanese studies57,58 demand analysis. Both studies used retrospective evaluation of stored still images, an approach that the IWGCO has found inadequate for reliable recognition of landmarks.62 The use of stored still images introduces another major flaw; Yamagishi et al.58 do not describe the criteria they used, but Akiyama et al.57 used the Prague Criteria, so how could they have arrived at a 43% prevalence of BE? Quite apart from the limitation of identifying landmarks in still images, the endoscopies themselves would not have been done in the way that is needed for adequate application of the Prague Criteria.32 It is almost certain that the position of the gastroesophageal junction was judged to be lower than its true position, because of effacement of the tops of gastric folds by the routine use of high levels of air distension during endoscopy in Japan. These Japanese studies appear so fatally flawed that their data must be considered invalid.

Prevalence estimates from clinical endoscopic series

Estimates of the prevalence of BE in reflux disease patients who are referred for endoscopy usually range from 10% to 15% in “Western” countries.2–4 These are mainly reports of endoscopically suspected BE and are probably under-estimates. In the past, especially in the USA, endoscopists have shrunk from recording the presence of “short” segments (usually less than 2 or even 3 cm in length), originally because of uncertainty whether these were really segments of esophageal columnar metaplasia and more recently, because of doubts about their clinical significance, despite acceptance that these were esophageal columnar metaplastic segments. Somewhat paternalistically, it seems to have been thought better not to open what was perceived as a proverbial can of worms. Yet most BE patients have metaplastic segments less than 3 cm long and EA does develop in metaplastic segments even shorter than 1 cm. Given the Prague criteria validation data,32 it is appropriate to identify at least all patients with segments of 1 cm or more so that their EA risk can be determined. If this is not done, we will continue to have no idea about how to advise and manage patients with segments less than 3 cm in extent. Part of this process should be to abandon the ambiguous description “short” and replace it with “C” and “M” values.32

Risk for development of dysplasia and adenocarcinoma

The now well-documented risk that BE carries for development of EA5 is the major concern of clinicians, because they are now expected manage this. Expectations and interest are being driven by the remarkable increase in the incidence of EA in relatively prosperous, mainly Caucasian, populations, albeit from a low base.2–4,53 The complex area of predictors of progression of BE to EA has been reviewed recently.63 Currently, in routine clinical practice, an individual patient's risk for EA is assessed only crudely by determining if dysplasia is absent or present and if it is present, whether it is of low or high grade. There is a pressing need for development of more exact and individualized approaches to assessment of EA risk.63,64 Tools that enable this will almost certainly be developed within the next decade (Fig. 2). There is now a sufficient knowledge base to allow development of a clinically valid risk-scoring system. There is the additional prospect that data on biomarkers currently being gathered31 will strengthen the predictive value of such novel instruments in due course. In the future, risk assessment instruments will probably not only improve the cost-effectiveness of surveillance, but also help to guide the use (or non-use) of endoscopic therapy or esophagectomy in the individual patient.

The sad pattern of presentation of esophageal adenocarcinoma does not undermine the logic of management of cancer risk in Barrett's esophagus

About 95% of patients with EA present at an advanced stage without prior diagnosis or surveillance of BE.2–4,64 This sad fact means that it is currently impossible for surveillance of BE to have any measurable impact on the overall mortality from EA. This reality is however no reason for out-of-hand rejection of measures designed to manage the risk for EA in patients in whom BE has been diagnosed.

Development of esophageal adenocarcinoma is relatively rare in Barrett's esophagus

The best available current estimate from endoscopic surveillance data is that 6.3 cases of EA develop during each 1000 person-years.5 Expression of risk in this way challenges the understanding of many clinicians, let alone patients.14,15 Put another way, 63 of 1000 BE patients will develop EA during 10 years of surveillance. This fairly small absolute risk is about 30 times that of the general population risk for EA3.

Confirmed low-grade dysplasia signals a major risk for esophageal adenocarcinoma

The key word in the title statement is “confirmed”. Vieth has pointed out that the frequency of diagnosis of low-grade dysplasia in nine reports on endoscopic screening and surveillance of BE ranges from 67.2% to 1.2% of patients!47 Such variation can only be explained by the problems with reporting of biopsies discussed previously. In the Netherlands, endoscopic surveillance of patients with expert pathologist-confirmed low-grade dysplasia found a cumulative risk of development of high-grade dysplasia or EA of 85% after 109 months.50 By contrast, in those whose diagnosis of low-grade dysplasia was reversed by expert review, surveillance over 107 months revealed high-grade dysplasia or EA in only 4.6%. Consistent with this, a UK study found that 27% of 34 patients with expert-confirmed low-grade dysplasia progressed to high-grade dysplasia or EA during 8 years of observation; only 5% of patients who did not develop low-grade dysplasia had such progression.65 It is notable that among pathologists at this expert UK center, the interobserver agreement for a diagnosis of low-grade dysplasia was only fair to good (Kappa values 0.42—0.70).65

When biopsies are reported as showing low-grade dysplasia, clinicians should have them reviewed by an expert because confirmation of true low-grade dysplasia identifies patients who are at such high risk for development of EA that management should be tailored to this fact.47,49,50 It appears however that there are very few pathologists who are consistently accurate in recognizing “true” low-grade dysplasia associated with high risk for EA.

There remains major uncertainty about the magnitude of the risk for progression of high-grade dysplasia to esophageal adenocarcinoma

In 1991, the lack of reliable EA risk estimates made management of high-grade dysplasia a medical variant of Russian roulette.1 Then, the only options were to do nothing or have an esophagectomy. Now there are risk estimates, but they still leave clinicians with an uncomfortably high level of uncertainty.

A recent systematic review appropriately concluded that there were only four technically acceptable reports66–69 on the EA risk among the 196 publications identified as reporting on high-grade dysplasia.70 These four studies have a combined experience of 1241 patient years in 236 patients. Figure 4 shows that the pooled weighted risk estimate for progression to EA was 65.8 per 1000 patient-years or, put another way, during one year of observation, a patient with high-grade dysplasia has a 6.6% risk for development of EA. This is a lower risk than most would believe is the case from their clinical experience.

image

Figure 4. Risk for development of esophageal adenocarcinoma in high-grade dysplasia for the four studies included as valid in the meta-analysis of Rastogi et al.,70 with the pooled risk rate for all studies combined. The range of reported risk varies by more than five times. Adapted from Rastogi et al.,70 with permission.

Download figure to PowerPoint

Figure 4 reveals the problem with a risk estimate of 65.8 per 1000 patient-years. The range of risk estimates for the four studies is almost five-fold! The Schnell study,68 which found an implausibly low risk of progression of high-grade dysplasia to EA of 22.7 per 1000 patient-years, accounts for 577 patient years, 46% of the pooled duration of observation of the four studies. The Schnell study also reported the extraordinary and highest-recorded incidence of low-grade dysplasia of 67.2%68 in Vieth's previously mentioned tabulation of studies.47 This author concludes that the only plausible explanation for the results of Schnell et al. is that what is categorized as high- (and low-) grade dysplasia in their institution differs substantially from other BE research centers.

The significant clinical implications of the discordant data on EA risk are discussed below in the section on management of high-grade dysplasia.

Most Barrett's esophagus patients die from causes other than esophageal adenocarcinoma

A review of reports on cause-specific mortality in BE patients undergoing surveillance found that only 9% of these patients die as a result of EA5. This could be interpreted as a triumph for surveillance, but the reality is that most BE patients die of other causes, most commonly cardiovascular, without development of high-grade dysplasia or EA, because they are, on average, elderly and have a high rate of comorbidities.5 Clinicians need to temper their choice of management options for EA risk in the light of this (Fig. 2).

Treatment of reflux disease in Barrett's esophagus

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

In 1990, it was noted that H2-receptor antagonists failed to heal esophagitis in a high proportion of BE patients and, in the first flush of PPI therapy, there were minimal data on the treatment of BE with these agents.1 Now, it is all very cut and dried, even though there is a lack of data from randomized controlled trials: if once-daily standard dose PPI fails to heal esophagitis or control reflux-induced symptoms in BE patients, addition of a second dose, preferably before the evening meal, will usually succeed, given that BE patients have high levels of nocturnal esophageal acid exposure.71 Further escalation of PPI dose is sometimes needed.

In the case of antireflux surgery, reports which appeared in the 1990s reached conflicting conclusions about the ability of fundoplication to control reflux adequately in BE patients. This led to trialing of some quite radical alternative approaches, such as vagotomy with partial gastrectomy and Roux-en-Y anastomosis.72 Happily, it is now clear that either open73 or laparoscopic fundoplication74 done by experts achieves excellent control of reflux in BE patients.

Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

This field is covered by a Cochrane review which this author finds particularly difficult to read.75 This is a confused but crucial area for clinicians. The confusion arises from unsubstantiated claims that antireflux surgery can prevent development of adenocarcinoma. Chemopreventive therapy is the most promising of several otherwise disappointing possibilities.

Why prevention of development of Barrett's esophagus is not an option

Early intervention would be needed and is not feasible

Superficially, prevention of development of BE is an attractive option for preventing the development of EA. The reality is that this strategy will probably never succeed, even if the factors that trigger the development of BE are fully understood. This is because if BE is not found at the first endoscopy, it develops only rarely in subsequent years.2,3 Therefore, prevention requires early and accurate identification of the population at risk before the usual time of presentation for a first endoscopy. Any intervention must be very safe and effective. This is such a tall order that it is highly unlikely to occur, except in the unlikely event of a paradigm-changing discovery about pathogenesis on a par with the discovery of H. pylori. Even if a potent preventive strategy were developed, it is unlikely to come anywhere near being cost-effective, given the relatively low overall risk for development of BE.

Prevention of Barrett's esophagus is a spurious indication for antireflux surgery

Despite the insight that BE rarely develops in reflux disease patients under observation, some vocal advocates claim that prevention of BE is one of the benefits of antireflux surgery. This claim springs from the unsubstantiated conviction that long-term treatment of reflux disease with PPI puts patients at risk for development of BE and ultimately cancer! This is either manipulative or a display of inadequate knowledge of the natural history of BE. There are simply no data which suggest that development of BE is a significant risk during PPI therapy, even after more than 20 years of increasingly wide use of PPI and endoscopy.

If prevention of BE is the primary reason for undergoing surgery, its use for this reason alone will cause significant net harm, since there is no logical expectation for any benefit with regard to adenocarcinoma risk. Inescapable harm arises from the cost, rare mortality, occasional major post-operative complications and significant morbidity from the symptoms caused by the mechanical effects of this surgery, even in centers of excellence.2,3

Acid suppression therapy has not been shown to protect against esophageal adenocarcinoma

There are no prospective, randomized controlled data on whether PPI therapy has any effect on risk for EA. Some cohort studies have used the questionable endpoint of development of dysplasia of any grade as a proxy for cancer. The major inaccuracy of the diagnosis of low-grade dysplasia by general and even specialist pathologists discussed above undermines the authority of these studies. The available observational data have been tabulated recently in a rather long, well-structured but conflicted letter to the Editor76 which initially concludes rather generously that “the available results strongly suggest that PPI therapy may prevent the progression of BE to neoplastic lesions”, but then states in the next sentence that “the available evidence is too limited to draw any definite conclusion.” This is the best “definite” conclusion that can be made from the current literature. Consequently, Fig. 2 does not list acid suppression as a risk-reducing intervention.

Long-term endoscopic cohort studies suggest that very long-term PPI therapy is associated with a minor reduction of extent of metaplasia and the appearance of more squamous islands. These changes are most unlikely to be associated with any useful reduction of cancer risk.

Esophageal pH monitoring studies have shown that many BE patients treated with once-daily PPI in the morning still have high levels of esophageal acid exposure, especially at night.71 These observations, and studies that show evidence of persistence of mucosal markers of ongoing esophageal mucosal injury during partial control of esophageal acid exposure, have sparked speculation that the lack of detectable effect on risk for EA from routine PPI therapy could be due to under-treatment. Consequently, it has been proposed that twice-daily PPI, given at a dose to “normalize” levels of acid reflux, might reduce EA risk.4 This is an optimistic speculation, in light of the negative data for a cancer-protective effect of antireflux surgery29,30 discussed immediately below. One study has found that twice-daily PPI has no impact on mucosal markers of injury.71 The large AspECT study, now in progress (see below), which has randomized patients to twice or once daily esomeprazole 40 mg, should provide definitive data on whether twice-daily PPI has any EA protective effect, compared to once-daily therapy.

Antireflux surgery has no detectable effect on adenocarcinoma risk

It is a biologically plausible hypothesis that the proven major impact of expert antireflux surgery on gastroesophageal reflux could reduce the risk for development of EA by transforming a highly aggressive esophageal luminal environment to one that is benign.4 Protection against EA has been claimed repeatedly by some surgeons as an established benefit of antireflux surgery on the basis that it makes “obvious sense”. This conviction was reinforced by an uncritical analysis of the data then available: this conclusion has been refuted by two later, more careful evaluations. The literature considered in these analyses, labors under many technical limitations.29,77 Despite this, the belief that antireflux surgery has a clinically valuable protective effect against EA in BE patients continues to be stated so vocally and frequently by a few that it has come to be widely believed. As a result, it is not unusual for belief in protection against EA to be a contributory reason, sometimes the only one, that leads patients to have antireflux surgery.78

The first and currently only high quality study of whether antireflux surgery protects against cancer was published in 2010.30 The records of all individuals having antireflux surgery in Sweden from 1965 to 2005 were evaluated and the incidence of EA in the 14 102 antireflux surgery patients was compared to controls. It is most unlikely that there will ever be any more authoritative data than this. The conclusion was: “Antireflux surgery cannot be considered to prevent the development of esophageal or cardia adenocarcinoma”.30

Believers in the cancer-protective effects of antireflux surgery claim that any negative data on whether surgery protects against esophageal adenocarcinoma are explained by the failure of substandard surgery to control reflux. This breath-taking conceit, which displays ignorance of the concept of intention to treat, is discussed and rejected by Lagergren et al.30

Given current knowledge, it is now completely inappropriate to imply to patients that any degree of protection against EA is a benefit of antireflux surgery (Fig. 2).

Chemopreventive therapy—the most promising option for low-cost reduction of risk for esophageal adenocarcinoma

Over the last decade quite a large number of observational studies have found consistently that aspirin and non steroidal anti-inflammatory drugs (NSAID) use is associated with around a 50% reduced risk for EA.79–81 A recent retrospective survey failed to find any benefit of aspirin use, but these data are likely to be subject to major confounding influences.82 Overall, the epidemiologic data provide strong logic for controlled, prospective trials of chemoprevention. Very large, long-term, randomized studies, with meticulous adjudication of histopathologic diagnoses are needed. A study which showed no benefit of celecoxib on risk of progression of grade of dysplasia or development of EA during only 48 weeks should be disregarded: though randomized, only 49 patients received celecoxib and 51 placebo.83 This was a study which never had a chance of testing the hypothesis that celecoxib protects against development of high-grade dysplasia or EA.

The relatively low absolute risk for EA in all patients with BE means that the number needed to treat to prevent one cancer is high. Consequently, any intervention must be inexpensive and safe. Low-dose aspirin, given with PPI, has the potential to measure up to this challenge. The strong possibility that aspirin is chemopreventive is being evaluated in the largest study ever undertaken in BE, the UK-based AspECT study, which satisfies the challenging criteria for a definitive study on chemoprevention. About 2500 males have been randomized to four treatment arms: once-daily esomeprazole 40 mg with or without low-dose aspirin and twice-daily esomeprazole 40 mg with or without low-dose aspirin. The design and power of this study is a much needed quantum leap in the quality of research that evaluates interventions in BE: it is sufficiently powered to allow use of the robust primary outcome measure of development of high-grade dysplasia or EA.

The first major data from the AspECT study will be available in 2012. Already, safety monitoring indicates that aspirin therapy combined with PPI has a low rate of serious adverse events (Prof J Jankowski, personal communication). If the chemopreventive effect of low-dose aspirin predicted by epidemiologic studies79,80 is confirmed, this is likely to become a widely recommended therapy for reduction of the cancer risk in BE patients (Fig. 2).

Low-dose aspirin and NSAIDs are not the only plausible candidates for chemoprevention of EA. Evidence of a chemopreventive effect of statins, suggested by cell biology studies,80 has also been found by a recent epidemiologic study.81 Several other options have also been proposed as worthy of investigation.80 Observational studies will need to give a consistently promising signal on the possible chemopreventive properties of a novel option before a definitive prospective, randomized intervention study is considered, because of the huge effort involved.

It is most unlikely that positive results from chemopreventive studies will alter recommendations for endoscopic surveillance in the near future, but in due course, such therapy might allow increases of intervals between surveillance endoscopies and so positively influence cost-effectiveness.

Mucosal ablation—too soon for routine use in non-dysplastic Barrett's esophagus

A meta-analysis has found that the risk for EA was 1.63 per 1000 patients-years in 6847 patients treated in uncontrolled studies with a range of mucosa-ablative therapies.84 This contrasts with an estimated risk of 5.98 per 1000 patient-years, determined in other studies of BE patients free of dysplasia who did not have ablative therapy.84 This impressive apparent risk reduction is potentially influenced by several confounders, but makes biological sense. The pros and cons of taking this aggressive approach in patients free of dysplasia are well reviewed by Sharma and colleagues85 who emphasize the need to weigh the risks and not insignificant cost of this intervention against the relatively low risk for EA in BE patients free of dysplasia.14 The number needed to treat to prevent one cancer, let alone one death from cancer is high, so the potential to harm is also high. In expert hands, the risks of ablation are relatively small. If mucosal ablation came to be widely practiced outside expert centers, its risks are likely to be greater in that setting. More randomized comparisons are needed on the long-term efficacy of the several options for ablation of metaplastic mucosa. Currently, mucosal ablation in patients with non-dysplastic BE should only be done within well-designed clinical trials.

Ablative therapy does not succeed in all patients and at least some methods are associated with rates of redevelopment of BE as high as 66% within 5 years of ablation85: reversion to BE at this rate will destroy any prospect of ablation providing net benefit. Ablative therapy continues to develop. The most recent option is radiofrequency ablation which in one study had a very high reported efficacy, with 97% of patients who had their non-dysplastic BE ablated being free of metaplasia 30 months post-therapy.86

Unfortunately, in the US, ablative therapy is already being widely used in patients with non-dysplastic BE by “competitive” clinicians before adequate definition of the risk/benefit balance.15 This reality should not distract researchers from the strong possibility that if ablative therapy permanently removes the need for ongoing endoscopic surveillance, it would be cost effective87 (Fig. 2). All should be revealed in the next 10 years!

Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

The choice of therapy for primary management of high-grade dysplasia seems such a politically charged topic that even very recent general reviews are disappointingly circumspect in their discussion of this.2–4

Major guidelines for BE management were published in 2005 (two), 2006 and 2008, and all list esophagectomy as an appropriate primary therapy for high-grade dysplasia.2,3 These guidelines would have taken at least one year to formulate and publish, so they rely on the literature available between 3 and 6 years ago. So much has been learnt since then about high-grade dysplasia and its management by endoscopic therapy that the recommendations of these guidelines for management of high-grade dysplasia are seriously outdated.

The diagnosis of high-grade dysplasia is not a catastrophe: it requires a measured response

Presence of high-grade dysplasia does not indicate that EA (even intramucosal) will develop in the immediate future (see above). If the worst case estimate of EA risk, that of Overholt is used69,70 (Fig. 4), the yearly risk of EA development in one patient is just over 10%. There is time to digest the following information.

Endoscopic therapy is highly efficacious in management of high-grade dysplasia, with negligible risk, minimal morbidity and at relatively low cost

Several expert centers have shown that endoscopic therapy is highly effective at removing and preventing recurrence of high-grade dysplasia for up to more than 5 years, with minimal risk and morbidity.89–94 Some of the data for high-grade dysplasia are a little difficult to tease out separately from outcomes of endoscopic therapy of intramucosal EA, but the excellent outcomes for EA attest to what endoscopic therapy can also achieve in high-grade dysplasia95 (Fig. 5). Endoscopic therapy is so safe and well-tolerated that in patients with a life-expectancy of several years, perhaps 5 years, the best option is not to wait and observe but to undertake mucosal resection directly on diagnosis of high-grade dysplasia. This approach will deal with the 3–12.7% risk of an undiscovered coexisting EA by examination of mucosal resection specimens.51,96 Targeted mucosal ablation may be needed to complete clearance of high-grade dysplasia. Use of endoscopic therapy does not preclude subsequent use of esophagectomy if examination of mucosal resection specimens reveals an EA that is unexpectedly penetrating into the submucosa.

image

Figure 5. Major outcomes related to expert esophagectomy or expert endoscopic therapy of esophageal adenocarcinoma, assessed as intramucosal at the time of primary therapy, from two published series.94,95 Comparisons between the non-randomized studies available are hazardous, because of many potential confounders. The follow-up duration of the endoscopic therapy study depicted is almost half that of the surgical data. Patient age (62.1, range 31–86 years) in the endoscopic therapy group was lower than in the surgical group (67.7 ± 1.4 SEM years). The 34% rate of significant treatment-related morbidity after esophagectomy was due to anastomotic leakage or stricturing, cardiopulmonary complications and feeding jejunostomy leakage. The reported outcomes of endoscopic therapy are so superior to those of esophagectomy, that they overcome concerns above possible confounding factors.

Download figure to PowerPoint

Endoscopic surveillance is essential after high-grade dysplasia has been ablated or resected. Further development of high-grade dysplasia or even EA occurs, but surveillance and re-treatment deal effectively with this risk.89–95 After an initial local mucosal resection of high-grade dysplasia, ablation or resection of the entire metaplastic mucosa is an effective option for dealing with the risks from an especially unstable metaplastic mucosa.92,93

Esophagectomy is an effective, hazardous, morbid and cost-ineffective treatment for high-grade dysplasia

The only advantage of esophagectomy for high-grade dysplasia is certainty that EA will not develop because the esophagus has been removed! This is a drastic remedy; total colectomy is not advocated for dysplastic adenomatous polyps. Perhaps the mind-set that still drives patients with high-grade dysplasia (and the surgeons they consult) to esophagectomy is determined by the lethality of the EA that presents at such an advanced stage outside surveillance programs. Yet, we now have ample evidence, consistent with experience in the colon, that surveillance-detected intramucosal EA, let alone high-grade dysplasia has very high cure rates when treated only by local therapy (Fig. 5).

The unacceptable price of esophagectomy (Fig. 5) compared to endoscopic therapy is firstly, well, its price! Management of the hazards of esophagectomy require major intensive care resources.98–99 Mortality from esophagectomy or just “scraping through” can be extremely expensive in terms of in-hospital costs. Secondly, death is a socially devastating and frequently costly problem, ranging from about 4–20%, depending on surgical and intensive-care expertise. Just over half of all esophagectomies are done in “low-volume” centers (< 7 cases per year) which have mortalities that range from 16.2% to 20.3%.50,98 The major morbidity associated with esophagectomy is the third major price, both immediate and long term.

Choice between endoscopic therapy and esophagectomy?—now a no-brainer!

Data on the efficacy of expert endoscopic therapy and the natural history of high-grade dysplasia make the use of esophagectomy as a treatment for high-grade dysplasia resemble taking a sledgehammer to crack open a coconut! Put another way, if this author had high-grade dysplasia, he would sell his beloved boat and wood-working equipment, even his house, if this were necessary for him to access expert endoscopic therapy for his high-grade dysplasia, in order to remain the owner of his esophagus and to avoid the consequences of what has now become an unnecessary esophagectomy.

The choice between endoscopic therapy and esophagectomy has not been, and never will be tested in a randomized clinical trial. Outcomes of endoscopic therapy for early EA, discussed below further reinforce the case for endoscopic management of high-grade dysplasia (Fig. 5). Gastroenterologists have been generally well-informed about the stark differences between the risk/benefit balance between endoscopic therapy and esophagectomy for high-grade dysplasia. A study of practice at the famous Massachusetts General Hospital has shown that between 2003 and 2007, if a patient with high-grade dysplasia or intramucosal EA was managed by a gastroenterologist, there was an 88% chance of being treated endoscopically: if a surgeon was in charge, the chance of being treated with esophagectomy was 86%.99 Perhaps practice has changed at the Massachusetts General since 2007, but today, large numbers of patients with high-grade dysplasia are still being treated with esophagectomy. Maybe this article will prevent some of these esophagectomies in the future.

Therapy of esophageal adenocarcinoma

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

Esophagectomy and endoscopic therapy play complementary roles in the management of EA. This is a specialized area which should remain in the hands of super-specialist surgeons or interventional endoscopists who are expert in management of BE patients.

An accurate endoscopic mucosal survey and tumor staging are essential to management

An endoscopy that reveals a large EA, with endoscopic ultrasound evidence of penetration into the esophageal mucosa cannot be managed endoscopically and any attempts at this will merely add unwarranted cost and complexity to management with esophagectomy.

When a small EA with only subtle surface mucosal changes is recognized, this must not distract the endoscopist from screening all of the other parts of the metaplastic mucosa for additional EA or high-grade dysplasia with suitable endoscopic equipment and mainly visually-guided biopsy.38 The topography of an early EA gives some guidance on whether it is just intramucosal,42 but endoscopic ultrasound is probably a waste of time in this setting.42,44 Endoscopic mucosal resection is the easily achieved gold-standard technique for staging of early EA. This gives reliable estimation of the chance of cancer cure by endoscopic therapy.45,46,89,91–95,99 A mucosal resection that has completely removed an EA is, of course highly effective therapy for EA, as well as a definitive staging procedure.

Endoscopic therapy of early esophageal adenocarcinoma gives very high cure rates

There does not appear to be a systematic review of the rapidly growing literature on the outcomes of endoscopic therapy of EA. Ell and colleagues, who have pioneered this area, report a 99% initial cure rate in 100 patients95 (Fig. 5). The one patient with “failure” of initial cure was eventually apparently cured by repeated endoscopic therapy, because he refused to have surgery. Patients were selected for endoscopic therapy on the basis of a range of detailed histopathologic criteria applied to endoscopic mucosal resection specimens. Recurrences or new development of EA that occurred in 11% of patients during a mean follow-up of just over 3 years were all treated successfully. The calculated 5-year survival rate was 98%.95 Other centers have reported somewhat lower, but still excellent cure rates.2–4,89,92–94,96,99 It is clear that processes for mucosal screening, patient selection, endoscopic resection technique and histopathologic assessment of biopsies and mucosal resection samples are all still being refined in many centers. Treatment of early EA with ablative therapy only is an inferior option to initial mucosal resection, since this approach does not allow accurate staging.

Effective options are available for managing cancer risk in remaining metaplastic mucosa

After successful endoscopic removal of an early EA, the significant risk of further EA in the metaplastic mucosa can be managed effectively by ongoing surveillance.89,91–96,99 Another approach though, is to resect or ablate the remaining metaplastic mucosa, after local resection of the EA.92,93,99 Vigorous, twice-daily PPI therapy is given to ensure that ablated areas heal with squamous mucosa. Of the ablative techniques, radio-frequency ablation appears the most promising in this setting.89,99 Only long-term surveillance in these patients will tell us whether complete ablation results in essentially complete reversal of the EA risk.

Esophagectomy is not necessary for treatment of intramucosal adenocarcinoma

The logic and data that show that esophagectomy is not an appropriate alternative to endoscopic therapy for high-grade dysplasia have equal validity to the treatment of intramucosal EA.

The place of endoscopic surveillance

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

Researchers who elect to evaluate the cost-effectiveness of endoscopic surveillance must have a masochistic streak, as the findings of completed studies are being constantly undermined by advances in the management of the EA risk in BE.87,100 Thus, by the time a cost-effectiveness study is designed, completed and published, the estimates and assumptions necessary for the study no longer reflect best current practice and its outcomes.

Studies of the cost-effectiveness of endoscopic screening and surveillance do however have an important role to play. They highlight how cost-inefficient surveillance is in many settings, especially in patients with non-dysplastic BE and therefore the need to improve this. Refusal to undertake endoscopic surveillance on grounds that it is not cost-effective is simply not an option for clinicians, given guidelines and patient expectations.2,3,14,15,101

Payers of healthcare costs are the only group that may be sufficiently empowered to act on cost-effectiveness data about BE surveillance by denial of reimbursement for this, on the grounds that, from a community perspective, it is an unjustified cost on the health system. Probably, few would be so bold.

Figure 2 shows graphically how the wide range of opportunities that has been reviewed in this article might contribute towards enhanced cost-effectiveness of endoscopic surveillance. Especially important is the need to develop tools for more accurate measurement of individual patient risk for development of an EA that is present long enough to cause major disability or death, by using variables such as gender, BMI, possibly extent of metaplasia, age, race, biomarkers, etc..31,63,64

Non-endoscopic screening or surveillance methods

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

Endoscopy is currently the only method available for management of the EA risk in BE. Screening and surveillance with such a costly and invasive technique is far from optimal, so it is important that efforts are made to find more simple options suited to use in screening and surveying the great majority of individuals at low absolute risk for development of EA.

Capsule sponge esophageal cytology—a near future screening method for Barrett's esophagus?

This method is not yet fully validated nor generally available. It appears to be the only relatively low-cost, non-endoscopic screening method for which there are clinical data. A cytology sponge is compressed and encased in a gelatin capsule attached to a string. The capsule, but not the end of the string is swallowed.102 After a few minutes in the stomach, the liberated sponge is dragged back up the esophagus. This procedure would seem to score high on any “yuck” scale, but it is reported to be well-tolerated and suitable for use in primary care.103 The problem of a “needle in a haystack” search of the recovered exfoliated mucosal cells for esophageal columnar metaplastic cells is solved not by cell morphological criteria, but by molecular biological identification of cells in which trefoil factor 3 is strongly expressed, a feature unique to esophageal columnar metaplasia. A pilot study in 96 controls and 36 BE patients found this test to have a sensitivity of 78% and a specificity of 94% for presence of BE.101 It is even possible that this technique, coupled with a panel of molecular markers, might be capable of screening for dysplasia or even EA. Watch this space!

Potential for non-esophageal markers to reduce need for surveillance

Many studies have found specific changes in the esophageal mucosa that are associated with subsequent development of EA. None of these has yet achieved the status of a validated biomarker,31 but there is still a good chance that this will occur (Fig. 2).

Study of esophageal mucosal factors imposes the practical and financial burden of getting “a bit” of the esophagus. The level of participation in screening programs for colon cancer is strongly influenced by the characteristics of the screening test. The ideal screening method for BE would be a blood test or buccal smear that might be applied to assess risk for BE in those with reflux-induced symptoms. There is hope that systemic104 or genetic markers27 of risk for BE could become well-enough characterized that a non-invasive BE screening method could be devised.

Acknowledgments

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References

The author is grateful to AstraZeneca for their very long-term and continual support of the International Working Group for the Classification of Oesophagitis, which has made possible the development of the Los Angeles Classification of reflux oesophagitis, the Prague C&M criteria and the Barrett's Oesophagus Related Neoplasia (BORN) project. The BORN project which will provide training on the diagnosis of early esophageal adenocarcinoma, is also facilitated by support from Olympus.

Esther Breed is thanked for her invaluable support with searching for and locating publications and for typing and making multiple revisions of the manuscript.

References

  1. Top of page
  2. Abstract
  3. Introduction—an historical perspective
  4. History, definitions and diagnostic criteria for Barrett's esophagus
  5. Pathogenesis of Barrett's esophagus and esophageal adenocarcinoma
  6. Diagnosis, grading of extent of Barrett's esophagus and assessment of mucosal histology
  7. Epidemiology of Barrett's esophagus
  8. Treatment of reflux disease in Barrett's esophagus
  9. Possible therapeutic strategies for prevention of development of esophageal adenocarcinoma
  10. Expert-confirmed low-grade dysplasia—a diagnosis that calls for action
  11. Expert-confirmed high-grade dysplasia—management by surgery or endoscopic therapy?
  12. Therapy of esophageal adenocarcinoma
  13. The place of endoscopic surveillance
  14. Non-endoscopic screening or surveillance methods
  15. Acknowledgments
  16. References
  • 1
    Dent J, Bremner CJ, Collen MJ, Haggitt RC, Spechler SJ. Working party report to the World Congresses of Gastroenterology, Sydney 1990. Barrett's oesophagus. J. Gastroenterol. Hepatol. 1991; 6: 122.
  • 2
    Shaheen NJ, Richter JE. Barrett's oesophagus. Lancet 2009; 373: 85061.
  • 3
    Sharma P. Barrett's esophagus. N. Engl. J. Med. 2009; 361: 254856.
  • 4
    Spechler SJ, Fitzgerald RC, Prasad GA, Wang KK. History, molecular mechanisms, and endoscopic treatment of Barrett's esophagus. Gastroenterology 2010; 198: 85469.
  • 5
    Sikkema M, De Jonge PJF, Steyerberg EW, Kuipers EJ. Risk of esophageal adenocarcinoma and mortality in patients with Barrett's esophagus: a systematic review and meta-analysis. Clin. Gastroenterol. Hepatol. 2010; 8: 23544.
  • 6
    Shaheen NJ, Crosby MA, Bozymski EM, Sandler RS. Is there publication bias in the reporting of cancer risk in Barrett's esophagus? Gastroenterology 2000; 119: 3338.
  • 7
    Lord RVN. Norman Barrett, “doyen of esophageal surgery”. Ann. Surg. 1999; 229: 42839.
  • 8
    Dent J. Review article: from 1906 to 2006—a century of major evolution of understanding of gastro-oesophageal reflux disease. Aliment. Pharmacol. Ther. 2006; 24: 126981.
  • 9
    Barrett NR. Chronic peptic ulcer of the oesophagus and “oesophagitis”. Br. J. Surg. 1950; 38: 17582.
  • 10
    Allison PR, Johnstone AS. The oesophagus lined with gastric mucous membrane. Thorax 1953; 8: 87101.
  • 11
    Barrett NR. The lower esophagus lined by columnar epithelium. Surgery 1957; 41: 88194.
  • 12
    Vakil N, van Zanten SV, Kahrilas P, Dent J, Jones R, Global Consensus Group. The Montréal definition and classification of gastroesophageal reflux disease: a global evidence-based consensus. Am. J. Gastroenterol. 2006; 101: 190020.
    Direct Link:
  • 13
    Takubo K, Vieth M, Aida J et al. Differences in the definitions used for esophageal and gastric diseases in different countries. Digestion 2009; 80: 24857.
  • 14
    Shaheen NJ. Risk and reason in Barrett's esophagus. Clin. Gastroenterol. Hepatol. 2010; 8: 2334.
  • 15
    Falk GW. Radiofrequency ablation of Barrett's esophagus, should everyone get it? Gastroenterology 2009; 136: 2399402.
  • 16
    Harrison R, Perry I, Haddadim W et al. Detection of intestinal metaplasia in Barrett's esophagus: an observational comparator study suggests the need for a minimum of eight biopsies. Am. J. Gastroenterol. 2007; 102: 115461.
    Direct Link:
  • 17
    Das D, Ishaq S, Harrison R et al. Management of Barrett's esophagus in the UK: overtreated and underbiopsied but improved by the introduction of a national randomized trial. Am. J. Gastroenterol. 2008; 103: 111.
    Direct Link:
  • 18
    Riddell RH, Odze RD. Definition of Barrett's esophagus: time for a rethink—is intestinal metaplasia dead? Am. J. Gastroenterol. 2009; 104: 25884.
  • 19
    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: 81624.
  • 20
    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: 6574.
  • 21
    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: 12714.
  • 22
    Lagergren J. Oesophageal cancer and gastro-oesophageal reflux: what is the relationship? Gut 2004; 53: 10645.
  • 23
    Sartori S, Nielson I, Indelli M, Trevisani L, Pazzi P, Grandi E. Barrett esophagus after chemotherapy with cyclophosphamide, methotrexate and 5-fluorouracil (CMF). An iatrogenic injury? Ann. Intern. Med. 1991; 114: 2101.
  • 24
    Spechler SJ, Schimmel EM, Dalton JW, Doos W, Trier JS. Barrett's epithelium complicating lye ingestion with sparing of the distal esophagus. Gastroenterology 1981; 81: 5803.
  • 25
    Csendes A, Braghetto I, Burdiles P, Korn O, Csendes P, Henríquez A. Very late results of esophagomyotomy for patients with achalasia: clinical, endoscopic, histologic, manometric, and acid reflux studies in 67 patients for a mean follow-up of 190 months. Ann. Surg. 2006; 243: 196203.
  • 26
    Maley CC. Open questions in oesophageal adenocarcinogenesis. Gut 2007; 56: 8978.
  • 27
    McElholm AR, McKnight A-J, Patterson CC et al. A population-based study of IGF axis polymorphisms and the esophageal inflammation, metaplasia, adenocarcinoma sequence. Gastroenterology 2010; 139: 20412.
  • 28
    Huo X, Zhang HY, Zhang X et al. Acid and bile salt-induced CDX2 expression differs in esophageal squamous cells from patients with and without Barrett's esophagus. Gastroenterology 2010; 139: 194203.
  • 29
    Chang EY, Morris CD, Seltman AK et al. The effect of antireflux surgery on esophageal carcinogenesis in patients with Barrett esophagus. Ann. Surg. 2007; 246: 1121.
  • 30
    Lagergren J, Ye W, Lagergren P, Lu Y. The risk of esophageal adenocarcinoma after antireflux surgery. Gastroenterology 2010; 138: 1297301.
  • 31
    Jankowski JA, Odze RD. Biomarkers in gastroenterology: between hope and hype comes histopathology. Am. J. Gastroenterol. 2009; 104: 10936.
  • 32
    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: 13929.
  • 33
    Lee YC, Cook MB, Bhatia S et al. Interobserver reliability in the endoscopic diagnosis and grading of Barrett's esophagus: an Asian multinational study. Endoscopy 2010; 42: 699704.
  • 34
    Reid BJ, Blount PL, Feng Z, Levine DS. Optimizing endoscopic biopsy detection of early cancers in Barrett's high-grade dysplasia. Am. J. Gastroenterol. 2000; 95: 308996.
    Direct Link:
  • 35
    Pech O. Declaration of bankruptcy for four-quadrant biopsies in Barrett's esophagus? Clin. Gastroenterol. Hepatol. 2009; 7: 6102.
  • 36
    Kariv R, Plesec TP, Goldblum JR et al. The Seattle protocol does not more reliably predict the detection of cancer at the time of esophagectomy than a less intensive surveillance protocol. Clin. Gastroenterol. Hepatol. 2009; 7: 6538.
  • 37
    Wolfsen HC, Crook JE, Krishna M et al. Prospective, controlled tandem endoscopy study of narrow band imaging for dysplasia detection in Barrett's esophagus. Gastroenterology 2008; 135: 2431.
  • 38
    Curvers WL, Bergman JJ. Multimodality imaging in Barrett's esophagus: looking longer, seeing better, and recognizing more. Gastroenterology 2008; 135: 297312.
  • 39
    Li M, Anastassiades CP, Joshi B et al. Affinity peptide for targeted detection of dysplasia in Barrett's esophagus. Gastroenterology 2010; 139: 147280. doi: 10.1053/j.gastro.2010.07.007.
  • 40
    Pohl H, Rösch T, Vieth M et al. Miniprobe confocal laser microscopy for the detection of invisible neoplasia in patients with Barrett's oesophagus. Gut 2008; 57: 164853.
  • 41
    Wallace MB, Sharma P, Lightdale C et al. Preliminary accuracy and interobserver agreement for the detection of intraepithelial neoplasia in Barrett's esophagus with probe-based confocal laser endomicroscopy. Gastrointest. Endosc. 2010; 72: 1924.
  • 42
    Pech O, Gossner L, Manner H et al. Prospective evaluation of the macroscopic types and location of early Barrett's neoplasia in 380 lesions. Endoscopy 2007; 39: 58893.
  • 43
    May A, Gossner L, Pech O et al. Local endoscopic therapy for intraepithelial high-grade neoplasia and early adenocarcinoma in Barrett's oesophagus: acute-phase and intermediate results of a new treatment approach. Eur. J. Gastroenterol. Hepatol. 2002; 14: 108591.
  • 44
    Pech O, May A, Günter E, Gossner L, Ell C. The impact of endoscopic ultrasound and computed tomography on the TNM staging of early cancer in Barrett's esophagus. Am. J. Gastroenterol. 2006; 101: 22239.
    Direct Link:
  • 45
    Badreddine RJ, Prasad GA, Lewis JT et al. Depth of submucosal invasion does not predict lymph node metastasis and survival of patients with esophageal carcinoma. Clin. Gastroenterol. Hepatol. 2010; 8: 24853.
  • 46
    Montgomery E, Canto MI. Management of high-grade dysplasia in patients with Barrett's esophagus. Clin. Gastroenterol. Hepatol. 2006; 4: 14349.
  • 47
    Vieth M. Low-grade dysplasia in Barrett's esophagus—an innocent bystander? Contra. Endoscopy 2007; 39: 6479.
  • 48
    Alikhan M, Rex D, Khan A, Rahmani E, Cummings O, Ulbright TM. Variable pathologic interpretation of columnar lined esophagus by general pathologists in community practice. Gastrointest. Endosc. 1999; 50: 236.
  • 49
    Montgomery E, Goldblum JR, Greenson JK et al. Dysplasia as a predictive marker for invasive carcinoma in Barrett esophagus: a follow-up study based on 138 cases from a diagnostic variability study. Hum. Pathol. 2001; 32: 37988.
  • 50
    Curvers WL, Ten Kate FJ, Krishnadath KK et al. Low-grade dysplasia in Barrett's esophagus: overdiagnosed and underestimated. Am. J. Gastroenterol. 2010; 105: 152330.
  • 51
    Zhu W, Appleman HD, Greenson JK et al. A histologically defined subset of high-grade dysplasia in Barrett mucosa in predictive of associated carcinoma. Am. J. Clin. Pathol. 2009; 132: 94100.
  • 52
    Montgomery E. Refining diagnostic criteria for high-grade dysplasia in Barrett esophagus. Am. J. Clin. Pathol. 2009; 132: 79.
  • 53
    Wong A, Fitzgerald RC. Epidemiologic risk factors for Barrett's esophagus and associated adenocarcinoma. Clin. Gastroenterol. Hepatol. 2005; 3: 110.
  • 54
    El-Serag H. The association between obesity and GERD: a review of the epidemiological evidence. Dig. Dis. Sci. 2008; 53: 230712.
  • 55
    Ronkainen J, Aro P, Storskrubb T et al. Prevalence of Barrett's esophagus in the general population: an endoscopic study. Gastroenterology 2005; 129: 182631.
  • 56
    Zou D, He J, Ma X et al. Epidemiology of symptom-defined gastroesophageal reflux disease and reflux esophagitis: the systematic investigation of gastrointestinal diseases in China (SILC). Scand. J. Gastroenterol. Published online: 19 Oct 2010.
  • 57
    Akiyama T, Inamori M, Iida H et al. Shape of Barrett's epithelium is associated with prevalence of erosive esophagitis. World J. Gastroenterol. 2010; 16: 4849.
  • 58
    Yamagishi H, Koike T, Ohara S et al. Clinical characteristics of gastroesophageal reflux disease in Japan. Hepatogastroenterology 2009; 56: 10324.
  • 59
    Peng S, Cui Y, Xiao Y et al. Prevalence of erosive esophagitis and Barrett's esophagus in the adult Chinese population. Endoscopy 2009; 41: 10117.
  • 60
    Tseng P, Lee Y, Chiu H et al. Prevalence and clinical characteristics of Barrett's esophagus in a Chinese general population. J. Clin. Gastroenterol. 2008; 42: 10749.
  • 61
    Park J, Kim J, Kim H et al. The prevalence of and risk factors for Barrett's esophagus in a Korean population. J. Clin. Gastroenterol. 2009; 43: 90714.
  • 62
    Armstrong DA, Bennett JR, Blum AL et al. The endoscopic assessment of esophagitis—a progress report on observer agreement. Gastroenterology 1996; 111: 8592.
  • 63
    Prasad GA, Bansal A, Sharma P, Wang KK. Predictors of profession in Barrett's esophagus: current knowledge and future directions. Am. J. Gastroenterol. 2010; 105: 1490502.
  • 64
    Reid BJ, Li X, Galipeau PC, Vaughan TL. Barrett's oesophagus and oesophageal adenocarcinoma: time for a new synthesis. Nat. Rev. Cancer 2010; 10: 87101.
  • 65
    Lim CH, Treanor D, Dixon MF, Axon ATR. Low-grade dysplasia in Barrett's esophagus has a high risk of progression. Endoscopy 2007; 39: 5817.
  • 66
    Weston AP, Sharma P, Topalovski M, Richards R, Cherian R, Dixon A. Long-term follow-up of Barrett's high-grade dysplasia. Am. J. Gastroenterol. 2000; 95: 188893.
    Direct Link:
  • 67
    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: 166976.
  • 68
    Schnell TG, Sontag SJ, Chejfec G et al. Long-term nonsurgical management of Barrett's esophagus with high-grade dysplasia. Gastroenterology 2001; 120: 160719.
  • 69
    Overholt BF, Lightdale CJ, Wang KK et al. Photodynamic therapy with porfimer sodium for ablation of high-grade dysplasia in Barrett's esophagus: international, partially blinded, randomized phase III trial. Gastrointest. Endosc. 2005; 62: 48898.
  • 70
    Rastogi A, Puli S, El-Serag HB, Bansal A, Wani S, Sharma P. Incidence of esophageal adenocarcinoma in patients with Barrett's esophagus and high-grade dysplasia: a meta-analysis. Gastrointest. Endosc. 2008; 67: 3948.
  • 71
    Abu-Sneineh A, Tam W, Schoeman M et al. Effect of high-dose esomeprazole on gastric and oesophageal acid exposure and molecular markers in Barrett's oesophagus. Aliment. Pharmacol. Ther. 2010; 32: 102330. Published online: 15 AUG 2010 DOI: 10.1111/j.1365-2036.2010.04428.x.
  • 72
    Csendes A, Burdiles P, Braghetto I, Korn O, Díaz JC, Rojas J. Early and late results of the acid suppression and duodenal diversion operation in patients with Barrett's esophagus: analysis of 210 cases. World J. Surg. 2002; 26: 56676.
  • 73
    Parrilla P, Martínez de Haro LF, Ortiz A, Munitiz V, Serrano A, Torres G. Barrett's esophagus without esophageal stricture does not increase the rate of failure of Nissen fundoplication. Ann. Surg. 2003; 237: 48893.
  • 74
    Attwood SE, Lundell L, Hatlebakk JG et al. Medical or surgical management of GERD patients with Barrett's esophagus: the LOTUS Trial 3-year experience. J. Gastrointest. Surg. 2008; 1: 164655.
  • 75
    Rees REJ, Lao-Sirieix P, Wong A, Fitzgerald R. Treatment for Barrett's oesophagus. Cochrane Database Syst. Rev. 2010; Issue 1, Art No CD 004060 DOI: 10.1002/14651858.CD004060.pub2.
  • 76
    Islami F, Kamangar F, Boffetta P. Use of proton pump inhibitors and risk of progression of Barrett's esophagus to neoplastic lesions. Am. J. Gastroenterol. 2009; 104: 2646.
  • 77
    Shaheen NJ. Does fundoplication change the risk of esophageal cancer in the setting of GERD? Am. J. Gastroenterol. 2005; 100: 100911.
    Direct Link:
  • 78
    Vakil N, Shaw M, Kirby R. Clinical effectiveness of laparoscopic fundoplication in a US community. Am. J. Med. 2003; 114: 15.
  • 79
    Corley DA, Kerlikowske K, Verma J, Buffler P. Protective association of aspirin/NSAIDs and esophageal cancer: a systematic review and meta-analysis. Gastroenterology 2003; 124: 4756.
  • 80
    Ilyas S, DeMars CJ, Buttar NS. Chemoprevention in Barrett's esophagus. J. Gastrointest. Cancer 2007; 38: 19.
  • 81
    Nguyen DM, Richardson P, El-Serag HB. Medications (NSAIDs, statins, proton pump inhibitors) and the risk of esophageal adenocarcinoma in patients with Barrett's esophagus. Gastroenterology 2010; 138: 22606.
  • 82
    Gatenby PA, Ramus JR, Caygill CP, Winslet MC, Watson A. Aspirin is not chemoprotective for Barrett's adenocarcinoma of the oesophagus in multicentre cohort. Eur. J. Cancer Prev. 2009; 18: 3814.
  • 83
    Heath EI, Canto MI, Piantadosi S et al. Secondary chemoprevention of Barrett's esophagus with celecoxib: results of a randomized trial. J. Natl. Cancer Inst. 2007; 99: 54557.
  • 84
    Wani S, Puli SR, Shaheen NJ et al. Am. J. Gastroenterol. 2009; 104: 50213.
  • 85
    Sharma P, Falk GW, Sampliner R, Spechler SJ, Wang K. Management of nondysplastic Barrett's esophagus: where are we now? Am. J. Gastroenterol. 2009; 104: 8058.
  • 86
    Fleischer DE, Overhlot BF, Sharma VK et al. Endoscopic ablation of Barrett's esophagus: a multicentre study with 2.5-year follow-up. Gastrointest. Endosc. 2008; 68: 86776.
  • 87
    Inadomi JM, Somsouk MA, Madanick RD, Thomas JP, Shaheen NJ. A cost-utility analysis of ablative therapy for Barrett's esophagus. Gastroenterology 2009; 136: 210114.
  • 88
    Shaheen NJ, Sharma P, Overholt BF et al. Radiofrequency ablation in Barrett's esophagus with dysplasia. N. Engl. J. Med. 2009; 360: 227788.
  • 89
    Triadafilopoulos G. Blitzkrieg for Barrett's esophagus containing early neoplasia. Clin. Gastroenterol. Hepatol. 2010; 8: 79.
  • 90
    Behrens A, May A, Gossner L et al. Curative treatment for high-grade intraepithelial neoplasia in Barrett's esophagus. Endoscopy 2005; 37: 9991005.
  • 91
    Pech O, Behrens A, May A et al. Long-term results and risk factor analysis for recurrence after curative endoscopic therapy in 349 patients with high-grade intraepithelial neoplasia and mucosal adenocarcinoma in Barrett's oesophagus. Gut 2008; 57: 12006.
  • 92
    Chennat J, Konda VJ, Ross AS et al. Complete Barrett's eradication endoscopic mucosal resection: an effective treatment modality for high-grade dysplasia and intramucosal carcinoma—an American single-center experience. Am. J. Gastroenterol. 2009; 104: 268492.
  • 93
    Moss A, Bourke MJ, Hourigan LF et al. Endoscopic resection for Barrett's high-grade dysplasia and early esophageal adenocarcinoma: an essential staging procedure with long-term therapeutic benefit. Am. J. Gastroenterol. 2010; 105: 127683.
  • 94
    Prasad GA, Wu TT, Wigle DA et al. Endoscopic and surgical treatment of mucosal (T1a) esophageal adenocarcinoma in Barrett's esophagus. Gastroenterology 2009; 137: 81523.
  • 95
    Ell C, May A, Pech O et al. Curative endoscopic resection of early esophageal adenocarcinomas (Barrett's cancer). Gastrointest. Endosc. 2007; 65: 310.
  • 96
    Konda VJ, Ross AS, Ferguson MK et al. Is the risk of concomitant invasive esophageal cancer in high-grade dysplasia in Barrett's esophagus overestimated? Clin. Gastroenterol. Hepatol. 2008; 6: 15964.
  • 97
    Lerut T, Nafteux PH, Moons J et al. Quality in the surgical treatment of cancer of the esophagus and gastroesophageal junction. Eur. J. Surg. Onccol. 2005; 31: 58794.
  • 98
    Bergman JJ. Endoscopic resection for treatment of mucosal Barrett's cancer: time to swing the pendulum. Gastrointest. Endosc. 2007; 65: 113.
  • 99
    Yachimski P, Nishioka NS, Richards E, Hur C. Treatment of Barrett's esophagus with high-grade dysplasia or cancer: predictor of surgical versus endoscopic therapy. Clin. Gastroenterol. Hepatol. 2008; 6: 120611.
  • 100
    Barbiere JM, Lyratzopoulos G. Cost-effectiveness of endoscopic screening followed by surveillance for Barrett's Esophagus: a review. Gastroenterology 2009; 137: 186776.
  • 101
    El-Serag HB, Naik AD. Surveillance in Barrett's esophagus: lessons from behavioral economics. Gastroenterology 2009; 137: 7635.
  • 102
    Lao-Sirieix P, Boussioutas A, Kadri SR et al. Non-endoscopic screening biomarkers for Barrett's oesophagus: from microarray analysis to the clinic. Gut 2009; 58: 14519.
  • 103
    Kadri SR, Lao-Siriex P, O'Donovan M et al. Acceptability and accuracy of a non-endoscopic screening test for Barrett's oesophagus in primary care: cohort study. BMJ 2010; 341: c4372. doi: 11.1136/bmj.c4732.
  • 104
    Rubenstein JH, Kao JY, Madanick RD et al. Association of adiponectin multimers with Barrett's oesophagus. Gut 2009; 58: 15839.