A guide for the endoscopic assessment of Barrett's esophagus. We welcome you to this educational DVD on the endoscopic assessment of Barrett's esophagus. This DVD will outline each of the areas, basics on Barrett's esophagus, standard endoscopic assessment, and advanced imaging modalities. We anticipate that this video will be instructive for fellows in training, endoscopists in practice, procedure room nurses and technicians, and anyone who wants a better understanding of Barrett's esophagus. Barrett's esophagus is a complication of chronic injury and inflammation in the setting of gastroesophageal reflux disease, and it is defined as the replacement of esophageal squamous lining by specialized intestinal mucosa. While it does not cause symptoms in and of itself, Barrett's esophagus is a known risk factor for esophageal adenocarcinoma. We've seen the incidence of esophageal cancer rising over the past several decades, as shown with the incidence curve in blue on this graph. The majority of patients who are diagnosed with esophageal cancer have a poor prognosis and die from this disease, as shown on the mortality curve in red, which follows the incidence curve closely above. The key to decreasing mortality from esophageal adenocarcinoma is early detection and opportunity to intervene. Barrett's associated dysplasia and early cancer are amenable to endoscopic therapies. Multiple modalities of endoscopic therapies may be utilized, such as endoscopic mucosal resection, radiofrequency ablation, also known as RFA, and cryotherapy. Endoscopic screening is not recommended for the general population. There is a strong recommendation with moderate level of evidence from the ACG to recommend male patients with chronic reflux symptoms of greater than five years duration and reflux symptoms that occur on at least a weekly basis who also have two or more risk factors for esophageal adenocarcinoma to undergo a screening endoscopy to identify the presence of Barrett's esophagus. These risk factors include 50 years of age or greater, Caucasian race, central obesity, cigarette smoking, and family history of esophageal cancer. In addition to reflux symptoms, long-term PPI use may be considered. Since women have a lower risk of developing esophageal cancer, screening may be considered on an individual basis in those patients that have multiple risk factors. The diagnosis of Barrett's esophagus requires two components, the endoscopic criteria and the histologic criteria. The endoscopic criteria are the presence of salmon-colored lining and sometimes seen as tongues projecting from the gastroesophageal junction in the tubular esophagus. When we see this, we should perform a minimum of eight biopsies in order to reduce sampling error to meet the histologic criteria. The histologic criteria are glomerular-lined epithelium with the presence of goblet cells. Oftentimes, we encounter the variable squamo-columnar junction in which we may see less than one centimeter of variability of the Z-line. If we biopsy this, we may have intestinal metaplasia, but this may be intestinal metaplasia of the cardia and not necessarily Barrett's esophagus. Pathologists are unable to distinguish between intestinal metaplasia of the cardia and Barrett's esophagus by histology alone. This is important to recognize because intestinal metaplasia of the cardia likely reflects pathology in the stomach and does not carry the same risk of esophageal cancer as Barrett's esophagus and would not benefit from endoscopic surveillance and treatment. For those patients who have been diagnosed with Barrett's esophagus without dysplasia, the recommendations from the ACG, the AGA, and the ASGE are to perform surveillance endoscopy procedures every three to five years to monitor for dysplasia. The British Society of Guidelines also recommends to use shorter intervals for those patients with long segments. There is a spectrum of pathology for Barrett's esophagus. First is the normal squamous epithelium. This may be replaced by Barrett's esophagus for specialized intestinal metaplasia. The neoplastic progression starts with low-grade dysplasia, which may progress to high-grade dysplasia. In dysplasia, we see features of cytologic atypia and architectural atypia. Intramucosal carcinoma is defined as when neoplastic involvement extends past the basement membrane into the lamina propria. When tumor extends past the muscularis mucosa into the submucosa, we have submucosal carcinoma, which is designated as invasive adenocarcinoma. There is high interobserver variability in the diagnosis of dysplasia among pathologists. Challenges include the confounding effects of inflammation and the possible presence of duplicated muscularis mucosa. Therefore, all cases of dysplasia should be confirmed with a second pathologist, ideally with expertise in gastrointestinal pathology. This histologic information provides us information on the risks associated with the disease. We have better understanding on the risk of lymph node metastasis, risk of harboring prevalent cancer, and risk of progression of disease. The implications for treatment are based on the risk of lymph node metastasis. The risk of lymph node metastasis in submucosal invasive disease is greater than 20%, but varies with depth of disease, and therefore, surgery and or systemic therapy may be indicated. On the other hand, the risk of lymph node metastasis in high-grade dysplasia is negligible, and accordingly, endoscopic therapy is now standard treatment. A meta-analysis demonstrated that the risk of lymph node metastasis in intramucosal carcinoma is on the order of 1-2%, and this warrants an endoscopic approach prior to esophagectomy, which carries significantly more mortality and morbidity. The surgical literature has reported among patients who underwent a prophylactic esophagectomy for the treatment of high-grade dysplasia in the setting of Barrett's esophagus, there is a risk of harboring occult invasive cancer with a pooled estimate of 40%. While pooled analysis does substantiate that, indeed, 40% of these cases with only high-grade dysplasia on biopsy prior to surgery did subsequently have cancer identified on the surgical specimen. The risk of submucosal invasive disease, which would not standardly be treated with endoscopy, was only 12% in an analytical review of these surgical cases. The majority of the cases of cancer were intramucosal carcinoma, which is, in most cases, amenable to endoscopic therapy. High-grade dysplasia is the best marker we have to identify who will go on to esophageal cancer. Among patients with high-grade dysplasia, the risk to progress to cancer is 6% per person per year. There is a high risk of prevalent cancer in the setting of a visible lesion. Oftentimes, low-grade dysplasia may be downgraded to nondysplastic Barrett's due to the initial diagnosis being over-called. True confirmed low-grade dysplasia has an overall risk of progression on the order of 0.4% per person per year. This risk of developing esophageal cancer among patients with nondysplastic Barrett's is estimated to be less than 0.3% per person per year, based on prospective cohort and population-based studies, and is much lower than previously estimated. These estimates on risk for disease progression, occult prevalent cancer, and cancer incidence help inform the management of Barrett's esophagus. Our goal is to perform and document a high-quality endoscopic assessment of Barrett's segments. We can use the five L's. These are landmarks, length, look carefully, lesions, and biopsy at multiple levels. The first L is landmarks. We want to identify and document esophageal landmarks for a complete assessment for Barrett's esophagus. By identifying the location of the diaphragmatic impression, the gastroesophageal junction, and the squamo-columnar junction, we can then appropriately assess and document the following states, normal, hiatal hernia, Barrett's esophagus, and Barrett's esophagus in the setting of a hiatal hernia. In this first example, we see that the diaphragmatic impression, the gastroesophageal junction, and the squamo-columnar junction are all at the same level, which is consistent with the normal state. In the second example, the gastroesophageal junction, demonstrated by the top of the gastric folds on the endoscopic view, and shown with the green line on the diagram, is displaced from the diaphragmatic impression. This is consistent with the presence of a hiatal hernia, which may vary with time and even within an endoscopic examination. In the third example, Barrett's esophagus is suspected, with the squamo-columnar junction, the pink line, being displaced from the gastroesophageal junction, with salmon-colored protrusions into the tubular esophagus. The final example demonstrates the displacement of both the squamo-columnar junction and the gastroesophageal junction, suggesting Barrett's esophagus in the setting of a hiatal hernia. The second L is length. We want to measure and document the segment length with the PROG-CNM classification. We know that long segments defined as 3 centimeters or greater have a greater risk of progression to cancer compared to short segments defined as 1 centimeter or greater, but less than 3 centimeters. A reliable and validated method to document length is to use the Prague classification in which we measure the top of the gastric folds to the circumferential extent of the Barrett segment and then to the maximal extent of the contiguous Barrett segment. We then designate the extents using letters C and M for the circumferential and maximal extents respectively. We also want to identify and document the extent of any areas of esophagitis. Erosive esophagitis may be documented with the LA classification system. In this example, we have a long segment of Barrett's esophagus with erosive esophagitis as noted here. We can take our measurements by first noting the diaphragmatic impression and then the top of the gastric folds. Then we note the circumferential extent and the maximal extent. In this example, we have a C3M5 segment. It is also important to note erosive esophagitis in terms of extent and classification. The third L is to look carefully for subtle lesions. We must train ourselves to recognize lesions and that is the best tool we can use in the assessment of the Barrett segment. Traditionally, we were suspicious when lesions protruded into the lumen. However, throughout the GI tract, we now know we must also look for more subtle lesions that may be flat or depressed. We do not detect what we see, but what we recognize. Another critical tool to use is a high resolution endoscope. These endoscopes offer 800,000 to 1,000,000 pixels and are superior to standard definition. If we focus down the lumen, we may miss more subtle lesions. If we examine the mucosa on FOS with tip deflection and attention to mucosal patterns, we may note subtle discoloration or pattern irregularities. The use of a soft transparent cap may also provide stable visual recognition of the lumen. We can also use a soft transparent cap The use of a soft transparent cap may also provide stable visualization while examining mucosa. This brings us to the fourth L, which is lesions. We want to identify, document, and target visible lesions as these lesions have a risk of harboring neoplasia as seen in these examples here. We can document the appearance of visible lesions with the Paris classification for superficial neoplasia. Protruding lesions that are pedunculated, 01P or 01S. Flat lesions that are slightly raised, 02A, completely flat. 02B or slightly depressed, 02C. And excavated lesions, 03. Those lesions that are protruding or depressed are at higher risk for harboring submucosal invasive disease than those that are 02A or 02B. We should target visible lesions with endoscopic mucosal resection. Biopsy specimens are subject to limited tissue, poor orientation, and crush artifact. Endoscopic resection provides a large and intact tissue specimen for more accurate diagnosis and staging. The diagnosis may change one-third to one-half the time with EMR compared to previous biopsies. Pathologists have less inter-observer variability with EMR compared to biopsies alone. In this case with a flat lesion, there was a question of foci of intramucosal carcinoma on endoscopic biopsies. However, crush artifact may lead to a crowded appearance making high-grade dysplasia versus intramucosal carcinoma a difficult diagnosis on biopsy. This patient underwent endoscopic resection for the diagnosis of intramucosal carcinoma by biopsy. The final pathology by endoscopic resection was downstaged and diagnosed as high-grade dysplasia. In the example of an ulcer, endoscopic biopsy again showed a foci of intramucosal carcinoma. However, endoscopic mucosal resection demonstrated invasive adenocarcinoma not previously detected on endoscopic biopsy. These examples portray how endoscopic resection is not only a therapeutic modality but is also an important diagnostic tool for staging lesions. Ablated modalities such as radiofrequency ablation can then be utilized to eradicate the remainder of the at-risk Barrett's epithelium. These ablated modalities, however, do not provide tissue acquisition for histologic assessment. Fifth L is levels. The standard Barrett's endoscopic assessment includes a biopsy protocol with multiple levels. After we have targeted visible lesions for tissue acquisition, either with biopsy or resection, we assess for occult or subtle neoplasia with a four-quadrant random biopsy at every one to two centimeter interval as recommended by the Seattle Protocol. In cases of dysplasia, one centimeter interval should be used. Advanced imaging modalities have been developed and investigated in order to improve the detection of dysplasia in Barrett's esophagus. These include but are not limited to chromoendoscopy, digital chromoendoscopy, confocal laser under microscopy, and optical coherence tomography-based technology. The goals of the technical performance of these modalities in surveillance for Barrett's esophagus are to improve the detection of dysplasia and to reduce the number of negative surveillance biopsies. While these technologies may provide an enhanced image, an important question is what are the thresholds for the diagnostic parameters for the incorporation into clinical practice? The ASGE Preservation and Incorporation of Valuable Endoscopic Innovations Initiative set PIVI criteria for imaging technologies in Barrett's esophagus. The recommended thresholds for sensitivity and negative predictive value for targeted biopsies are 90% and 98% respectively. Chromoendoscopy is an enhanced imaging technique that utilizes dye on the tissue. In Barrett's esophagus, methylene blue, indigo carmine, and acetic acid have been investigated. Of the agents studied in ASGE meta-analysis, acetic acid was the only chromoendoscopy agent to meet PIVI criteria for sensitivity, specificity, and negative predictive value. Acetic acid is a contrast agent and upon contact with the tissue temporarily enhances mucosal pattern with a whitish change. Note the regular pit pattern seen on the left and the irregular distorted pattern seen on the right, which is concerning for neoplasia. In this patient with a long segment Barrett's esophagus, a flat lesion is noted as shown by the arrow. Acetic acid is safe, readily available, and inexpensive. Acetic acid is used to enhance the mucosal features. Appreciate how the regular mucosal pattern is at the GE junction and the distal esophagus. Now look at the distorted mucosal pattern and loss of pit pattern noted at the lesion previously shown, which was diagnosed by EMR as high-grade dysplasia. Digital chromoendoscopy alters the image to provide an enhanced image. Narrowband imaging is the most commonly used and investigated type of digital chromoendoscopy. Narrowband imaging filters light to the blue end of the spectrum so that hemoglobin absorbs the light and the mucosal patterns and vascular patterns are enhanced. Digital chromoendoscopy is widely available, less cumbersome than the use of staining dyes, and may be activated with the push of a button. This modality is also available in combination with the near-focus magnification as seen here. Several criteria have been proposed to investigate for narrowband imaging. The Barrett's International NBI Group developed an internally validated classification system, which essentially allows for classification of mucosal and vascular patterns into regular and irregular. These examples demonstrate the regular appearing patterns where mucosal pit patterns are all round, ridged or phyllous, or tubular, and the vascular patterns follow the mucosal ridges or have long, branching networks. Contrast this to the irregular appearing patterns which have absent or irregular pit patterns, or vascular patterns that are focally or diffusely distributed and do not follow mucosal ridges. Narrowband imaging does meet PIVI criteria for sensitivity, specificity, and negative predictive value thresholds. Notice, in this case, how narrowband imaging can be used to clearly differentiate the borders of a lesion. In this patient with high-grade dysplasia, a lesion is noted by white light endoscopy in the upper right-hand corner of the left eye. In this patient with high-grade dysplasia, a lesion is noted by white light endoscopy in the proximal esophagus. Narrowband imaging can be used to delineate the borders of the lesion. See how the mucosal pattern and the vascular pattern transitioned to the irregular patterns here. This lesion may now be targeted for vocal EMR, with the borders having been delineated by narrowband imaging. In this patient referred with low-grade dysplasia in the setting of a long segment Barrett's esophagus, a careful examination with white light demonstrated possible mucosal irregularities as shown in the circle. Narrow band imaging allowed more visualization of the mucosal pattern, and this lesion was targeted for EMR and confirmed high-grade dysplasia and low-grade dysplasia on pathology. The concept of confocal laser microscopy is shown in this diagram. We can see here a low-power laser going through an objective and emanating back through the same optical path through the detector. If we focus the light through a pinhole, this light is then focused through an objective to a single point in a defined microscopic field of view, and out-of-focus light is rejected. This allows for the plane of illumination to be the same as the plane of detection. Instead of using a single pinhole, the probe-based confocal laser endomicroscopy system uses thousands of optical fibers to create a pinhole effect. All detected signals from the illuminated spot are recaptured and measured. All signals are then processed to create a grayscale image, which represents one focal plane within the examined specimen. Image of scanned regions can be constructed from successive points and create a sharp image of tissue at that depth. Classification systems have been proposed for the interpretation of nondysplastic barrets versus dysplastic barrets in barret's esophagus. Here we see the Miami Criteria representation of normal lining in the squamous epithelium. We can appreciate the flat, scale-like cells and the interspersed blood vessels that are the papillary projections in the squamous lining. Notice the ONVAS imaging. In nondysplastic barrets, we appreciate a regular villiform architecture with regular epithelial borders and regular pulmonary line cells. We can appreciate the interspersed dark mucin-containing goblet cells. In a continuum of high-grade dysplasia to adenocarcinoma, we see the villiform architecture initially disrupted to disorganize and to eventual loss of the architecture. Classificational criteria have been proposed to distinguish between dysplastic and nondysplastic tissue. Images associated with dysplasia are likely to have two or more of the following features. Epithelial surface appears saw-toothed, goblet cells not easily identified, glands are not equidistant, glands are unequal in size and shape, cells are enlarged, and cells are irregular and not equidistant from one another. Imaging can be performed during an endoscopic procedure where the probe is inserted into the working channel of the endoscope. The probe is placed in contact with the tissue at light pressure. Imaging does require the use of a fluorophore, which in the case of the GI tract is intravenously administered fluorescein sodium. Under the administration of IV fluorescein, images can be obtained and interpreted in real time during acquisition or during SINAE review with the use of a trackball to go back to the desired frame or with recording the sequence and images to be reviewed offline at a later time. In the previous patient with the long segment Barrett's esophagus, there was an area that did not have readily apparent lesions seen on either white light endoscopy or narrowband imaging. CLE, however, showed some more worrisome features that could be suspicious for dysplasia, including cells not being regular or equidistant, glands varying in size and shape, an irregular epithelial stripe, glands not being equidistant, and a lack of goblet cells in this area. The EMR specimen demonstrated high-grade dysplasia. Volumetric laser endomicroscopy is a technology that is analogous to ultrasound, only light instead of sound penetrates the tissue and is measured. In this diagram, we see pulses of light going towards the tissue, and the back-reflected light is measured. The depth of penetration varies with time delay. The laser scanning probe is housed inside a balloon catheter to enhance centering in the esophagus. This catheter can be placed in the working channel of any standard endoscope. The images captured provide a three-dimensional dataset over a span of 6 centimeters to 3 millimeters of depth and 7 microns of resolution. Frames can be selected on a touchscreen interface and reviewed in a transverse zoom or longitudinal zoom for a detailed assessment offline. This image demonstrates characteristic squamous epithelium with layered architecture, with the squamous, muscularis mucosa, laminar propria, submucosa, and muscularis propria clearly identified. Here we have the gastrocardia with the appearance of rugae and pits. Next we have nondysplastic Barrett's epithelium, where we see a loss of layered architecture and the presence of glands. Imaging with volumetric laser endomicroscopy is safe and feasible. The difference between squamous, gastrocardia, and Barrett's is identifiable. Features associated with dysplasia have been proposed. These include loss of layering, surface intensity, which is higher at the surface than the subsurface, atypical irregular glands or ducts, and effacement. Criteria and scoring systems are being developed, refined, and validated for characterizing dysplasia. This case is a patient with a long-segment Barrett's esophagus with high-grade dysplasia, referred for evaluation and management. On white light examination, there was not a visible lesion identified. This is the narrowband imaging of the gastroesophageal junction and the more proximal esophagus. The VLE probe was placed into the esophagus. Volumetric laser endomicroscopy cross-sectional imaging was obtained in a 3-dimensional dataset. Upon image review, an area was identified that was suspicious. There was loss of layering characteristic of Barrett's esophagus, as well as septated irregular glands and darker irregular surface compared to subsurface. This area was identified, and here it is shown with narrowband imaging with near focus and targeted for endoscopic mucosal resection. The histology of this EMR specimen demonstrated high-grade dysplasia. The remainder of the resections and biopsies had no evidence of dysplasia. Here we see on higher power the cribriform glands and high-grade dysplasia next to the VLE image. The risk of cancer progression in Barrett's esophagus is a critical context in the endoscopic assessment and management. The standard endoscopic assessment includes identification and documentation of landmark, length, and visible lesions. Careful examination with high-resolution endoscopy and a protocol for assessment of occult or subtle neoplasia is recommended. Advanced imaging modalities may be utilized to improve detection of dysplasia.

Barrett's esophagus

esophageal adenocarcinoma

early detection

dysplasia

endoscopic therapies

advanced imaging modalities

surveillance endoscopy