Updated Guidelines 2008 for the Diagnosis, Surveillance and Therapy of Barrett's Esophagus

PREAMBLE The guidelines for the diagnosis, surveillance and therapy of Barrett's esophagus were originally published by the American College of Gastroenterology in 1998 and updated in 2002. These and other guidelines undergo periodic review. Significant advances have occurred in the area of Barrett's esophagus over the past four years leading to another revision of the prior guidelines. These advances include the potential use of esophageal capsule endoscopy for the diagnosis and screening of Barrett's esophagus, data regarding the outcome of low-grade dysplasia, the treatment of high-grade dysplasia using photodynamic therapy, and the development of new ablation techniques such as radiofrequency ablation. These guidelines are intended to be applied by physicians who see Barrett's esophagus patients and are intended to indicate a preferred, but certainly not the only, acceptable approach. Physicians need to choose the course best suited to the individual patient and to the variables that exist at the time of decision making. The guidelines are for adult patients with the diagnosis of Barrett's esophagus, as defined herein. Both these and the original guidelines were developed under auspices of the American College of Gastroenterology and the Practice Parameters Committee and approved by the Board of Trustees. The world literature was reviewed extensively for the original guidelines and once again reviewed using the National Library of Medicine database. Search terms used included Barrett's esophagus, esophageal neoplasm, esophagus, intestinal metaplasia, esophageal diseases, and adenocarcinoma, all appropriate studies and any additional ones found in reference to these papers were obtained and reviewed. Evidence was available from a hierarchy of trials and randomized controlled trials were given the greatest weight. Abstracts presented at national and international meetings were only used when unique data from ongoing trials were presented. When scientific data were lacking, recommendations are based on expert opinion. The recommendations made are based on the level of evidence found. Grade A recommendations imply that there is consistent level 1 evidence (randomized controlled trials), Grade B indicates that the evidence would be level 2 or 3 which are cohort studies or case control studies. Grade C recommendations are based on level 4 studies meaning case series or poor quality cohort studies, and Grade D recommendations are based on level 5 evidence meaning expert opinion. SIGNIFICANCE OF BARRETT'S ESOPHAGUS Barrett's esophagus continues to be increasingly recognized in the United States and is believed to be the major risk factor for the development of esophageal adenocarcinoma. The incidence of adenocarcinoma of the esophagus continues to rise rapidly. The rate of rise is alarming and is widespread in Western countries. In a review by the epidemiologists of the National Cancer Institute of cancer incidences normalized to the year 1975, esophageal adenocarcinoma incidence rates were found to outpace even those of melanoma, breast cancer and prostate cancer in terms of the rapidity of rise (1). These epidemiologists also found there was no concomitant decrease in diagnoses of gastric cancers or more proximal cancers, making a classification change unlikely to be responsible for this increase in adenocarcinoma. In the Danish Cancer Registry, adenocarcinoma incidence rates actually decrease in patients older than 85 (14.14/100,00 (80–84 yr) decreasing to 7.2/100,000 (85+ yr) unlike squamous cancer rates suggesting that this rise in adenocarcinoma incidence may be truly a recent phenomenon as evidenced by this age cohort effect (2). DEFINITION OF BARRETT'S ESOPHAGUS Barrett's esophagus is a change in the distal esophageal epithelium of any length that can be recognized as columnar type mucosa at endoscopy and is confirmed to have intestinal metaplasia by biopsy of the tubular esophagus. (Grade B recommendation). This working definition of Barrett's esophagus has changed little over the last 10 years. A recent "critical review of the diagnosis" of Barrett's esophagus concluded that "the working definition of BE is displacement of the squamocolumnar junction proximal to the gastroesophageal junction" and "endoscopy with multiple systematic biopsies is needed to establish the diagnosis of Barrett's esophagus" (3). This definition does not distinguish between short and long segment Barrett's esophagus and implies that only columnar lined esophagus should be biopsied. Although intestinal metaplasia is not specifically mentioned in this definition, clearly the reason to do multiple biopsies in the columnar appearing esophagus is to identify the presence of intestinal metaplasia, the premalignant lesion for esophageal adenocarcinoma (EAC). The vast majority of adenocarcinomas of the esophagus are accompanied by intestinal metaplasia in multiple cohort studies (4–8) and many adenocarcinomas of the esophagogastric junction are also associated with esophageal intestinal metaplasia (9–11). The incidence of adenocarcinoma of the esophagus has continued to rise in the United States, at least until the year 2002 (12). Supporting the primary role of BE as the premalignant lesion for EAC is the unmasking of underlying BE by chemotherapy of adenocarcinoma of the distal esophagus. A retrospective study reviewed 79 patients with locally advanced EAC who had preoperative chemotherapy and had restaging endoscopy and biopsy prior to resection. Pre-therapy endoscopy showed BE in 75%, whereas 97% had documented BE on post-chemotherapy biopsy or in the resected specimen (13). This suggests that the cancer overgrows the fertile field of BE so that at presentation of the patient with EAC, BE may no longer be detectable. Esophagitis might also mask Barrett's esophagus. In a recent study of 172 patients with erosive esophagitis, a full 12% were found to have Barrett's metaplasia after healing of the esophagitis (14). There is not universal agreement on the inclusion of intestinal metaplasia as a criterion for BE. The British Society of Gastroenterology has excluded the need for IM from the diagnosis of BE (15). It is well recognized that the yield of IM decreases as the segment of columnar lining shortens and fewer biopsies are taken. Repeat endoscopy and biopsy are often necessary to establish the presence of IM (16, 17). In patients with >1cm of columnar lined esophagus at endoscopy, multiple biopsies may be necessary to confidently detect intestinal metaplasia. Based on a recent retrospective study, eight biopsies may provide an adequate assessment of the presence of intestinal metaplasia (18). The issue becomes when to label a patient as having BE and having an increased risk for EAC compared to someone lacking BE. Because of the implication of the label of BE in the United States for obtaining health insurance and the increased cost of life insurance in the United States (19), it seems appropriate to establish the presence of IM before committing the patient to the diagnosis of BE and to surveillance endoscopy. There are no data on the risk of EAC in columnar lined esophagus lacking IM. Another new development in the endoscopic standardization of Barrett's esophagus is the Prague classification system of circumferential (CM) and maximal length (M). This system identifies the landmarks of the squamocolumnar junction, the gastroesophageal junction, the extent of circumferential columnar lining and the most proximal extension of the columnar mucosa excluding islands to determine the length of Barrett's esophagus. Twenty-nine endoscopists scored 29 videos with centimeter intervals marked on the image (20). The reliability coefficients (RC) for C 0.95, M 0.94, the gastroesophageal junction 0.88 and the location of the hiatus 0.85 were excellent. The overall RC for the endoscopic recognition of BE ≥1cm was 0.72. However, for less than 1cm of columnar lining the coefficient was only 0.22. In an era of growing endoscopic therapy for neoplastic BE, this standardization is important. Unfortunately, proximal islands of columnar lining and ultra-short BE 40 (21), heartburn (21–23), long duration GERD symptoms (more than 13 years) (23), and male gender (22). Yet the only consistent correlation in most studies was heartburn and the sensitivity was poor. With the nation's increasing obesity problems, it is not surprising that increased body mass index is correlated with Barrett's esophagus, particularly visceral adiposity characterized by CT scan of the abdomen (24). The emerging data on the potential mechanistic role of cytokines from increasing visceral fat will bear watching. The epidemiology of EAC in the United States identifies risk factors of male gender and Caucasian ethnicity: the annual incidence of EAC in Caucasian men is 3.6/100,000 compared to 0.8 in African American men and 0.3 in Caucasian women (12). The precise magnitude of risk for gender, ethnicity and age are not defined. Esophageal capsule endoscopy is a new technique that has the potential to provide a noninvasive diagnosis of suspected BE, i.e. a columnar lined esophagus. Early studies of small numbers of patients showing high sensitivity have been followed by data sets in abstract form documenting substantially lower sensitivity (25, 26). Although intriguing, this technique cannot be recommended in the screening setting at this time (Grade B Recommendation). It is anticipated that the cost of the capsule and its accuracy will be barriers to lowering the threshold for screening for BE. A more definitive estimate of the population prevalence of BE −1.6% - provides evidence of asymptomatic BE. Forty-four percent of the BE patients from a random sample of adults in 2 communities in Sweden lacked "troublesome heartburn and/or regurgitation over the past 3 months" (27). The inability to distinguish these patients' poses a major problem in developing an effective screening strategy for BE based upon symptoms. There are no current risk factors recognized to identify asymptomatic patients with BE. Such identification will be necessary before screening can be expected to effectively detect the majority of patients with BE. The natural history of asymptomatic BE is undefined. In summary, screening for Barrett's esophagus in the general population cannot be recommended at this time. (Grade B recommendation) The use of screening in selective populations at higher risk remains to be established (Grade D recommendation) and therefore should be individualized. SURVEILLANCE OF BARRETT'S ESOPHAGUS The grade of dysplasia determines the appropriate surveillance interval. Any grade of dysplasia by histology should be confirmed by an expert pathologist. Surveillance endoscopy remains controversial because of the lack of randomized trials supporting its value. Critical analysis of the literature does suggest a survival advantage of endoscopic surveillance. Multiple retrospective studies have been published, all of which indicate that survival is statistically enhanced if the cancers are detected by endoscopic surveillance rather than presenting with symptoms (Table 1). In a California community-based population, surveillance detected cancer had lower staging with better survival (28). A larger SEER/Medicare database documented that an EGD 1 year prior to the diagnosis of EAC was associated with earlier stage and improved survival (29).Table 1: Retrospective Surgical Series of Survival for EAC Based on Surveillance StatusSurveillance is practiced by the vast majority of endoscopists in the US (30, 31). The strongest rationale for early case detection of EAC is the poor 5 year survival of EAC of 13% even with contemporary therapy (32). A patient with documented BE needs to be assessed as a candidate for surveillance. It is recommended that patients be advised of the benefits and risks of surveillance endoscopy. Consideration for beginning a surveillance program should include age, likelihood of survival over the next five years, patient's understanding of the process and its limitations for detection of cancer, and the willingness of the patient to adhere to the recommendations (Grade B Recommendation). Surveillance endoscopy should be performed in patients whose reflux symptoms are controlled with proton pump inhibitor therapy. The goal is healing the esophagitis to reduce the likelihood of the inflammatory process interfering with the visual recognition of BE (14) and contributing to cellular changes confusing the reading of dysplasia. Four quadrant biopsies every 2cm of the Barrett's mucosa sample only a small fraction of the lining but offer the possibility of recognizing dysplasia. Ideally the biopsies from a given segment of Barrett's esophagus should be submitted to pathology in a separate container to enable the focusing of subsequent biopsies on the area if dysplasia is identified. Cost effectiveness studies are needed to evaluate this approach. Even if the initial two endoscopies within one year lack dysplasia, there is no guarantee of the subsequent lack of neoplasia, but may allow an interval of three years for surveillance (Table 2). A combined cohort of BE patients documented that half of patients who developed HGD/EAC had no dysplasia on their first two endoscopies (33).Table 2: Dysplasia Grade and Surveillance IntervalThe finding of low grade dysplasia (LGD) warrants a follow-up endoscopy within six months to ensure that no higher grade of dysplasia is present in the esophagus. If none is found, then yearly endoscopy is warranted until no dysplasia is present on two consecutive annual endoscopies. LGD should be confirmed by an expert GI pathologist because of the problem of reading variability (34). When two pathologists agree on the diagnosis of LGD, the patient has a greater likelihood of neoplastic progression percent of biopsies the recognition of LGD will be (20). of patients with LGD had no dysplasia after a follow-up of 4 years. The finding of high grade dysplasia in mucosa should to by an expert GI pathologist and a subsequent endoscopy within three with should undergo endoscopic resection. Although the natural history of is there is a five year risk of EAC excluding in the first It is because of the high risk of cancers that these patients are often as if cancer is with endoscopic CT and even have been performed there is not evidence to their Patients with confirmed high grade dysplasia, even if should be regarding their or would use as a threshold for or surveillance. who to have their dysplasia on surveillance should be to the highest of dysplasia found. This is based upon the problem of on subsequent of intestinal metaplasia mucosa can also with short segments of columnar so the patient should undergo periodic surveillance. If therapy has been patients should be followed and in the area of prior Barrett's mucosa at intervals appropriate for their prior grade of dysplasia until there is of ablation is documented on at least three consecutive endoscopies. (Grade D recommendation) surveillance is recommended Barrett's mucosa has been to recommendations regarding these intervals are not made given the of data of intestinal metaplasia but case series have established that the phenomenon does In summary, the surveillance of Barrett's esophagus does have evidence suggesting The more advanced the in terms of dysplasia, the more surveillance is However, using evidence of dysplasia as the primary to establish surveillance is There are with and need for endoscopies which this an that will need Surveillance is recommended but is a Grade C as long controlled studies are not OF grade dysplasia expert pathologist and more endoscopy and grade dysplasia also by an expert pathologist and a threshold for A more biopsy is necessary to the presence of concomitant adenocarcinoma. Any such as or is best assessed with endoscopic for a more and of of patients with high grade dysplasia is on endoscopic and and the patient's age, and is no longer the necessary treatment to have that for high-grade dysplasia the of four quadrant biopsies should be every 1 because larger intervals to a greater rates of cancer In any within the Barrett's if high-grade dysplasia has been found, should undergo endoscopic to adequate for more has been to be associated with a higher of and with to endoscopic may be these are The use of large has been in the setting of high-grade dysplasia, to biopsy have not been in terms of changes in patient The endoscopic technique to be used to yield is a which should the mucosa in to the biopsy Endoscopic has also been used surveillance of Barrett's esophagus in the that increased to sample the might to better diagnoses are as to additional can be obtained from However, the use of new such as in may be in increasing the clinical of ablation therapy has also been to decrease the risk of development of cancer within Barrett's esophagus. This is in with which to be a The of has not been established However, all studies on ablation therapy have been in with at least and most often proton pump inhibitor therapy. therapy has been the only therapy in a randomized control to decrease cancer risk in Barrett's esophagus In this study, patients were randomized to photodynamic therapy or with the primary of therapy using and was to decrease the risk of by but not the development of cancer after at least months of follow The therapy was also to high-grade dysplasia in of patients of patients in the control also high-grade dysplasia These were if high-grade dysplasia at any subsequent endoscopy. ablation techniques were originally for the treatment of Barrett's esophagus lacking dysplasia. The initial were that The of ablation was first with these ablation has been with or which to have based upon recent small randomized trials at high has been in case series to be to high-grade dysplasia and even small cancers, follow-up is not available has been used to low-grade dysplasia and rates of the Barrett's mucosa are in the with multiple of the of the have been in small of patients followed over short of time. therapy with an with has been in It is in high-grade dysplasia and early EAC in case series It does have of and even patient ablation using a based system has been to be of in of Barrett's esophagus in months after of treatment a radiofrequency on the has treatment of with this This was to the of the esophagus with high radiofrequency and esophageal have been Endoscopic of has also been to Barrett's esophagus, there is data its Surgical has been a of therapy for Barrett's esophagus with high grade dysplasia based upon that endoscopic surveillance may not detect early cancers in to of patients and the for prior to development of cancer may be the of EAC at in patients with at biopsy has been as low as recent studies have that the risk of cancer in the setting of is low at if there is no evidence of cancers detected in the presence of prior high grade dysplasia are early stage This has to changes in the is performed in these can be performed with techniques that the use of and However, the invasiveness of the one large series of patients the major rates time in and time of to be to that from which the of the esophagus the mucosa and with has also been in to decrease the after This has been to but has not been by the because of the need for the Patients need to be to a higher for the best A recent analysis of the literature has there needs to be at least a year at an to decrease mortality to or less A recent retrospective study comparing the mortality of patients with high-grade dysplasia with photodynamic therapy and endoscopic compared with found between the two at months of patients in had an esophageal cancer In summary, high-grade dysplasia is associated with a risk of cancer needs to be with of endoscopic ablation therapy, and presented to the patient based on their for these and the available to provide the current it as if surveillance with endoscopic techniques a of or may in retrospective cohort studies from expert The of which of these be and will on the available in the patient's the patient's and the (Grade B recommendation). BARRETT'S ESOPHAGUS Barrett's esophagus has been the of new It is not surprising the esophagus is using and the of mucosa to be is There have been to image Barrett's esophagus. The most available technique is a of the to two major and which are actually more by in the mucosa and These the to the mucosa better in with a high This has been the has been or A can be performed after image and has been by another The is based on with of by the Both of these can be applied to Barrett's esophagus In one study of patients with Barrett's esophagus with of had high grade dysplasia, the sensitivity of detection for a was with a of However, studies regarding the in of these has not been has also been used in to of dysplasia in Barrett's esophagus. This to detect from cellular in the esophagus. of dysplasia do not have as as and This may be more for screening larger of In Barrett's esophagus, one study has found that was for of high grade dysplasia in patients but had a rate have been used to image the esophagus with to the mucosa of of intestinal metaplasia but will not if there is high grade dysplasia or cancer The by which is applied and the of performed prior to of the this technique have had and studies have not found a advantage to in to random four quadrant biopsies in detection of dysplasia such as and have also been to the detection of in Barrett's esophagus in with high endoscopy There is in these it is the identification of will be in clinical The can the other techniques have been developed that small mucosa that might be on these These include and which can the mucosa and actually image cellular studies are in in Barrett's esophagus. in patients had an accuracy of for detection of which in a more to but using to also has for the detection of intestinal metaplasia at the gastric junction prior studies have not been in dysplasia can the from the mucosa and its to determine the of dysplasia that is that can such as and have been combined to allow improved of the mucosa the present of these is to in an and Although there is not evidence at this time to the use of these systems on a clinical BARRETT'S ESOPHAGUS Multiple have been but have actually been There is in the use of such as and in biopsy in cancer as well as of of such as and In recent studies that of and as well as demographic of the patients and BE length are of cancer or is for clinical There is a large cohort of patients that has been followed with the in the This has been using of that have been by to a of Based on these studies, there is no risk of cancer development for five years if there is no evidence of increased than or However, if was there was an increased risk of cancer whereas evidence of increased risk However, these have been to clinical because of the number of biopsies in the needed to In the in has at of as a using to detect of of and these are of cancer with a increased risk of cancer if of is detected However, these techniques have only been applied to that have been of these in a study is needed before it can be recommended for In a recent an of from patients who had developed cancer compared to case who had not found that of three and in their once again to predict cancer These be on which is an advantage over the mentioned However, these studies have only been on patient and have not been applied in a large that have been over time include of of and Although multiple have been to be in small of none of these has been in studies. The for the detection of GERD patients who will to BE would be noninvasive i.e. and or The to risk patients with BE would be noninvasive and the focusing of surveillance endoscopy on this high risk for EAC. This would identify patients with BE who will to EAC early for even the appropriate therapy. A low risk also be which might not cost effective surveillance would be this that can be performed on a clinical for widespread use are not BARRETT'S ESOPHAGUS a in the stage of esophageal adenocarcinoma by Barrett's esophagus seems Unfortunately, evidence that any treatment cancer and more cancer in this setting is The best evidence for any with that have been in multiple studies to be associated with a risk of cancer with an of interval This risk has also been with the that such as and were also with Unfortunately, in a randomized not its patient was not more effective than in patients with BE and dysplasia in the of the change of the of biopsies with dysplasia studies have risk of cancer in given trials are the use of and low and high proton pump inhibitor therapy in Barrett's esophagus but these will years to from two retrospective cohort studies suggest that therapy the likelihood of developing dysplasia This provides a rationale to even asymptomatic BE patients with The of therapy as a of cancer has not been documented can be made to use these as BARRETT'S ESOPHAGUS patients with Barrett's esophagus, the goal of with such as the proton pump is to control reflux symptoms. symptoms can be controlled in most patients with proton pump inhibitor therapy. a may be necessary in a subgroup of Retrospective studies have a decrease in development of dysplasia in patients with or proton pump have that of esophageal may decrease of there are no data that the use of high therapy to or the development of EAC. Patients who are for may These include patients lacking major and whose reflux symptoms are controlled with therapy. are with a rate at 5 years The vast majority of data do not provide that EAC detection of It is anticipated that in the short may available that identify Barrett's mucosa based on high or A randomized impact of surveillance endoscopy. A randomized controlled of surveillance is needed to determine the of this recognition of techniques are available that can distinguish of dysplasia. These from and to such as and or more of these will definition of risk of dysplasia. in the of endoscopic ablation of the most recent radiofrequency ablation is is beginning clinical trials and older are more photodynamic therapy with the development of new of the and duration of the surveillance after endoscopic ablation therapy of a to risk BE There are many potential but clinical trials that their This will change given the many for