- Email: [email protected]
Some observations on Barrett esophagus and associated dysplasia
Annals of Diagnostic Pathology 37 (2018) 75–82
Contents lists available at ScienceDirect
Annals of Diagnostic Pathology journal homepage: www.elsevier.com/locate/anndiagpath
Review Article
Some observations on Barrett esophagus and associated dysplasia a,⁎
b
c
T
a
Elizabeth Montgomery , Christina A. Arnold , Dora Lam-Himlin , Kevan Salimian , Kevin Watersd a
Department of Pathology, Johns Hopkins University, United States of America Department of Pathology, Ohio State University, United States of America Department of Pathology, Mayo Clinic Scottsdale, United States of America d Department of Pathology, Cedars Sinai Health System, United States of America b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Barrett esophagus Dysplasia Endoscopic mucosal resections Duplicated muscularis mucosae
Biopsy samples from esophageal columnar metaplasia and dysplasia are commonly encountered in Western pathology practice and knowing a few pitfalls can save both pathologists and patients a great deal of anxiety. Herein we discuss criteria for Barrett esophagus, evaluation of dysplasia, and some pitfalls in reviewing endoscopic mucosal resections. Also included is a summary of suggested follow-up for patients with Barrett esophagus.
Barrett esophagus is not defined uniformly worldwide. In the United Kingdom and Japan, the definition differs from that in the United States (US). The British [1], American Gastroenterological Association (AGA) [2], and American College of Gastroenterologists (ACG) [3] criteria appear in Table 1. The key difference between the US and British guidelines is that goblet cells are part of the US definition whereas they are not required in the British definition. However, the 2016 definition from the American College of Gastroenterologists - columnar epithelium with goblet cells extending ≥1 cm above the top of the gastric folds [3] - makes the life of the pathologist challenging. Whereas we frequently have a good idea about the length of a segment of columnar epithelium in question, in other instances, the only information that we receive for a sample is that it is labeled “esophagus”. Of course, if we receive a sample labeled “esophagus, 40 cm” and there is intestinal metaplasia and we have a second sample that is labeled “esophagus, 34 cm” and there is intestinal metaplasia, it is clear that the affected segment of lesion measures at least 1 cm. In fact, the gastroenterology colleagues who prepared the recommendations even went so far as to ask that our endoscopy colleagues refrain from taking biopsies of the gastroesophageal junction in the absence of visible alterations. However, there seems to be very little compliance with the latter suggestion. The American College of Gastroenterology applied the term “specialized intestinal metaplasia of the esophagogastric junction” for lesions with goblet cells that do not meet the 1 cm length requirement [3]. To address the length issue, prepared notes can be useful for situations for which 1) intestinal metaplasia is present without
⁎
knowledge of the segment length or 2) samples labeled “gastroesophageal junction”/GEJ with intestinal metaplasia. 1. Sample notes: Barrett mucosa Situation 1. Barrett mucosa, negative for dysplasia. See note. Note: The above diagnosis of Barrett esophagus is made due to presence of goblet cells (intestinal metaplasia) with the assumption that the biopsies were obtained from columnar mucosa in the distal esophagus and the mucosal irregularity extends at least 1 cm above the top of the gastric folds as per 2016 American College of Gastroenterology (ACG) guidelines. Reference: Shaheen NJ, Falk GW, Iyer PG, Gerson LB; American College of Gastroenterology. ACG Clinical Guideline: Diagnosis and Management of Barrett's Esophagus. Am J Gastroenterol. 2016 Jan;111(1):30–50. Situation 2. Cardiac mucosa with intestinal metaplasia. See note. Note: This biopsy shows gastric-type mucosa with scattered goblet cells. The diagnosis in this case depends on the location of this biopsy. If this biopsy was taken from the tubular esophagus and the mucosal irregularity extends at least 1 cm above the top of the gastric folds, it shows Barrett mucosa of the distinctive type. If this biopsy was taken from the gastric cardia, it shows intestinal metaplasia of the gastric cardia. Reference: Shaheen NJ, Falk GW, Iyer PG, Gerson LB; American College of Gastroenterology. ACG Clinical Guideline: Diagnosis and
Corresponding author at: 410 North Broadway, Weinberg Building Room 2242, Baltimore, MD 21231, United States of America. E-mail address: [email protected] (E. Montgomery).
https://doi.org/10.1016/j.anndiagpath.2018.09.013 Received 23 September 2018; Accepted 26 September 2018 1092-9134/ © 2018 Elsevier Inc. All rights reserved.
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Table 1 Definition of Barrett esophagus. Society
Definition
British (and Japanese) definition of Barrett mucosa 2014 American Gastroenterological Association definition of Barrett mucosa 2011 American College of Gastroenterologists' definition of Barrett mucosa 2016
- Columnar epithelium with or without goblet cells extending ≥1 cm above the gastric folds [1] - Columnar epithelium in the esophagus that contains goblet cells – no length requirement [2] - Columnar epithelium with goblet cells extending ≥1 cm above the top of the gastric folds [3]
[11,12]. These authors studied long-segment Barrett esophagus cases (> 3 cm) and noted superficial and deep CK7 immunoreactivity in the intestinalized mucosa, with only superficial CK20 staining in the intestinalized zones. In contrast, distal gastric intestinal metaplasia showed patchy, superficial, and deep CK20 staining in areas of incomplete intestinal metaplasia; strong, superficial, and deep CK20 staining in areas of complete intestinal metaplasia; and patchy or absent CK7 staining in either type of gastric intestinal metaplasia. Other studies have not confirmed these findings and CK7/20 immunolabeling has fallen out of favor. Other immunostains that have been studied include mucin core (MUC) polypeptides, which seem to be of little practical value. However, the key markers are MUC5 (gastric foveolar mucin), MUC6 (cardiac glands, antral glands, Brunner glands), and MUC2 (goblet cells). CDX2 staining has also been used to label areas of intestinal metaplasia [13] and some have noted that cases lacking goblet cells express these markers and believe that this supports the need to eliminate the requirement for goblet cells [14]. Hepatocyte antigen (Hepar-1, Carbamoyl Phosphate Synthetase 1) also marks intestinalization in the absence of goblet cells [15]. However, in daily practice it is more practical to simply search for goblet cells. None of the immunostains offers added value over H&E stains in detecting goblet cells [10].
Management of Barrett's Esophagus. Am J Gastroenterol. 2016 Jan;111(1):30–50. There are some observers in the US who have suggested eliminating the requirement for goblet cells in diagnosing Barrett mucosa since some esophageal adenocarcinomas arise in the absence of intestinal metaplasia. As an example, in one study, the authors reviewed endoscopic mucosal resection samples from a cohort of German patients and found adjoining intestinal metaplasia in association with less than half with early cancers [4]. However, these authors made no attempt to learn if the patients had separate samples with intestinal metaplasia. In two subsequent studies from the US West Coast, high grade columnar epithelial dysplasia and carcinomas were essentially always accompanied by intestinal metaplasia [5,6]. Similar results were found in an East Coast study [7] such that we would endorse retaining the requirement for goblet cells, a view not shared by all [8]. Indeed, there are some examples of esophageal adenocarcinomas that are unassociated with intestinal metaplasia but these are not numerous in our Western population. Eliminating the requirement for goblet cells would even further reduce an already unfavorable cost effectiveness of surveillance for esophageal adenocarcinoma. Endoscopically, Barrett mucosa consists of velvety “salmon colored” epithelium that can as extend as “tongues” above the gastric folds (Fig. 1). When an area of Barrett mucosa is surrounded by squamous mucosa, the appearance is referred to as an “island”. Our endoscopy colleagues use the Prague system to describe the extent of Barrett mucosa. In this system, the distance of the circumferential length of Barrett mucosa is recorded (“C”) and the maximum length is recorded as well (“M”) [9]. This method allows standardization of endoscopy reports and when these data are provided to pathologists, they afford some confidence in our diagnoses. In general, no special stains are needed to confirm the presence of goblet cells in esophageal biopsies. This topic was comprehensively reviewed by Panarelli and Yantiss, who concluded that neither histochemical nor immunohistochemical stains add value over H&E stains since they produce false positives [10]. In the past, the concept of using CK7/CK20 stains to separate esophageal intestinal metaplasia from gastric cardiac intestinal metaplasia was introduced by Ormsby et al.
2. Grading Barrett dysplasia The categories that are used to interpret biopsies [16]: Negative for dysplasia Indefinite for dysplasia Low grade dysplasia High grade dysplasia Adenocarcinoma Assessing Barrett biopsies is usually straightforward since most cases are nondysplastic but it is well known that observer variation can be an issue [17]. We have tightened our criteria in the last few years using a novel but very simple method to assess cell polarity (the relationship of cells one to another). We have also suggested that the combined number of cases diagnosed as indefinite for dysplasia, low grade dysplasia, and high grade dysplasia should not exceed 10% [18]. Of course, those clinics specializing in dysplasia would be expected to have a higher percentage of dysplasia cases. In evaluating Barrett mucosa, essentially the sample should be assessed for surface maturation and glandular crowding, its cytologic features, and whether inflammation is an obscuring factor before making a diagnosis. 2.1. Barrett esophagus, negative for dysplasia Nondysplastic Barrett mucosa should show surface maturation, which can be a challenge to confirm in suboptimally embedded samples. Minor nuclear alterations in the bases of the metaplastic pits are acceptable. However, noting the polarity of the epithelial cells and how they are arranged with respect to one another makes assessment of Barrett mucosa relatively easy. We have assessed Barrett mucosa easily in most cases by paying attention to “the four lines” [18]. Finding “the four lines” indicates preserved polarity of epithelial cells in both gastric
Fig. 1. This is an endocropic image from a patient with Barrett esophagus. The gastric folds are seen at the left of the image and a tongue of metaplastic epithelium is present. Note that the squamous epithelium at the right has a greyish pearly appearance. 76
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 2. Barrett mucosa with no dysplasia. Although scattered goblet cells are present, the area with the arrows indicating “the lines” lacks them and has an appearance like that of gastric cardiac surface epithelium. The top line is formed by the apical neutral mucin cap and the second by the base of that cap. The third line consists of the eosinophilic cytoplasm and the fourth by the layer of nuclei.
Fig. 4. Barrett esophagus indefinite for dysplasia. Inflammation and suboptimal sectioning are both factors. The surface nuclei are hyperchromatic and surface “lines” are somewhat obscured, particularly at the upper left of the image but the nuclei are not particularly enlarged.
reduce any inflammation followed by repeat sampling [3]. Of course no matter which images we show to illustrate our conception of cases that we would regard as indefinite for dysplasia, some colleagues will dismiss the illustrated changes as reactive and others will be concerned that dysplasia has been overlooked. That is the entire point of this category. It merits follow-up but the patient should not receive definitive treatment (mucosal ablation therapy) until the findings are clarified. The indefinite category tends to reflect lesions that are inflamed, situated in the bases of the pits with surface maturation, or for which artifact (diathermy) impede interpretation (Figs. 4–5). This indefinite category should be used in only a small percentage of cases (up to 3–5% but this is high) In fact, some colleagues essentially never use this category.
2.3. Low grade dysplasia Fig. 3. Barrett mucosa with no dysplasia. The lines described in Fig. 2 are indicated. In this area, the lines are interrupted by goblet cells.
Low grade dysplasia should be unequivocally neoplastic (adenomalike) and involve the surface epithelium. It is important not to overdiagnose low grade dysplasia as current guidelines endorse mucosal
cardiac mucosa and in Barrett mucosa (with goblet cells). Cardiac mucosa and Barrett mucosa are annotated in Figs. 2–3 to demonstrate the four lines at high magnification, but they can be identified at 4×. The top line comprises the apical mucin cap of neutral mucin typical of Barrett mucosa (with incomplete intestinal metaplasia) and cardiac mucosa. The second is composed of the bases of the mucin caps. The third reflects the aligned cytoplasm and the last (fourth) line is formed by the row of nuclei. Attention to these patterns makes it easy to dismiss many cases as reactive. 2.2. Barrett esophagus indefinite for dysplasia The concept of uncertainty in diagnosis was introduced in a 1983 study of epithelial changes in inflammatory bowel disease and was a radical change [19] since it was incomprehensible to our clinical colleagues that sometimes we are unable to discern whether epithelial changes are dysplastic/neoplastic or not. The term was later applied to columnar epithelial changes in Barrett mucosa [16,17] and, until recently, gastroenterology societies did not attempt to address it in guidelines. This category is presently regarded as a holding diagnosis until the findings in the patient are clarified by a course of treatment to
Fig. 5. Barrett esophagus indefinite for dysplasia. This is a higher magnification of the lesion seen in Fig. 4. 77
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 6. Low grade dysplasia. The surface nuclei hyperchromatic (adenoma-like) and there is an abrupt transition (arrow) between the zone in question and the remaining tissue.
Fig. 8. High grade dysplasia. Note the jumbled appearance of the nuclei at the surface. The nuclei are quite hyperchromatic and there is plenty of nuclear stratification.
ablation for low grade dysplasia and mucosal ablation confers a risk for stricture formation. In prospectively evaluated patients presenting to a clinic, the incidence of low grade dysplasia should be about 2–3% but not > 5% [20]. The nuclei in low grade dysplasia are larger than those of nondysplastic Barrett mucosa and generally there is little inflammation in samples confidently diagnosed with low grade dysplasia. A very helpful clue is an abrupt transition between the dysplastic zone and adjoining zones that are clearly not dysplastic (Figs. 6–7). The surface lines that characterize nondysplastic Barrett mucosa are effaced. Some examples of low-grade dysplasia can resemble colorectal tubular adenomas. Low grade dysplasia demonstrates loss of the four lines (an overall indication of altered cell architecture and cell polarity) but maintained nuclear polarity such that the long axes of the nuclei should remain more or less perpendicular to the basement membrane. We would note that there is some subjectivity in separating low grade dysplasia and high grade dysplasia but this is less important than it was in the past; all dysplasia is currently managed by endoscopic ablation [1,3]. Staining for TP53 is sometimes useful in confirming an interpretation of low grade (as well as of high grade) dysplasia - a few darkly stained nuclei on the surface can be identified and the basal glands label.
Most dysplasia cases show intestinal differentiation in that the epithelium is stratified and contains goblet cells - similar to the appearance of colorectal adenomas, but not all do [21]. 2.4. High grade dysplasia High grade dysplasia is usually easy to recognize. The area appears hyperchromatic at low magnification, standing out from any nondysplastic mucosa in the sample. The alterations are usually detected in the surface epithelium. Prominent nucleoli are not typical of high grade dysplasia. Usually there is ample lamina propria between the glands. Many nuclei both in the pits and at the surface have lost their relationship to the basement membrane and have rounded up and are arranged in a jumbled configuration (Fig. 8). Most cases of high grade dysplasia show some nuclear elongation and stratification similar to that in colorectal adenomas. However, some cases show an unusual pattern of small tubules, each lined by a monolayer of hyperchromatic nuclei (Fig. 9). Since the monolayer lacks nuclear stratification, it has been referred to as a “nonadenomatous” form of high grade dysplasia by some [22] whereas others have considered this pattern as evidence for gastric differentiation in high grade
Fig. 9. High grade dysplasia. The indicated glands (arrows) that have a monolayer of round hyperchromatic nuclei are dysplastic (high grade). Some examples of high dysplasia show only this type of glands.
Fig. 7. Low grade dysplasia. This is a higher magnification of the process seen in Fig. 6 showing the abrupt transition (arrow). 78
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 10. High grade dysplasia. TP53 immunolabeling labels the nuclei intensely.
Fig. 12. High grade dysplasia. This is a TP53 stain from the lesion seen in Fig. 11. There is light labeling in the nuclei of the normal cardiac glands (wild type pattern) and some of the lymphoid cells but complete absence of labeling in the dysplastic glands, suggesting biallelic inactivation of the TP53 gene.
dysplasia [23,24]. Immunolabeling for TP53 can be helpful to confirm an impression of high grade dysplasia when there is doubt. If the histologic features are already classic, performing immunolabeling adds no value, but about 85–90% of the time, if immunolabeling is added, strong labeling confirms the diagnosis (Fig. 10). TP53 is a tumor suppressor protein normally active during cell division and has a short half-life. Since TP53 is a normal molecule, some labeling is to be expected in the proliferative compartment of the mucosa. Therefore there is nuclear labeling in the basal layer of the squamous epithelium and in the pits of the stomach, the proliferative compartments of these tissues. When the TP53 gene is mutated, this results in a TP53 protein with a long half-life, so the protein accumulates in the nuclei of the cells and can be detected by immunolabeling. There is also a subset of high grade dysplasia cases in which biallelic loss of the TP53 gene results in complete absence of the protein, termed the “null pattern”. This can be exploited for diagnosis just as well as the strongly reactive pattern (Figs. 11–12). Finally, there are cases of both low grade and high grade that display either gastric foveolar differentiation or differentiation similar to that seen in pyloric gland adenomas This type of dysplasia can arise in a background of intestinal metaplasia but it still has gastric type differentiation [21]. The nuclei in this pattern of dysplasia are paler than those seen in conventional type dysplasia and this form of dysplasia is
Fig. 13. Foveolar pattern low grade dysplasia. Although there are a few goblet cells, the low grade dysplasia seen at the surface is characterized by enlarged nuclei and loss of “the lines”.
Fig. 14. Foveolar pattern low grade dysplasia. This is a higher magnification of the lesion shown in Fig. 13.
Fig. 11. High grade dysplasia. This image allows comparison between the dysplastic nuclei and the nuclei of a few cardiac glands. 79
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 15. Basal pattern dysplasia (basal crypt dysplasia). Note the pocket at the lower left, which jumps out because the nuclei are strikingly hyperchromatic. There is not unequivocal surface involvement. Fig. 17. Intramucosal carcinoma. There are glands with luminal necrosis and the one in the upper center of the lesion grows parallel rather than perpendicular to the surface.
diagnosed by attention to nuclear enlargement (low grade dysplasia) (Figs. 13–14); and mutilayering and loss of nuclear polarity in high grade dysplasia.
tissue beneath the mucosa is not present for evaluation, but a good estimate can be made on mucosal biopsies alone! Because of this, we diagnose intramucosal carcinoma (invasion into the lamina propria) on mucosal samples because we know that this interpretation will prompt an endoscopic mucosal reaction we can then better assess for deeper invasion. However, some institutions diagnose “adenocarcinoma” with a note that the patient should be evaluated clinically to determine whether the lesion can be managed endoscopically. Intramucosal carcinoma denotes invasion into the mucosa and thus is staged as T1a [27], which differs from the colorectal staging scheme for which lamina propria invasion is staged as Tis because there is minimal lymphatic access in the lamina propria of the colon. Invasion into the submucosa by esophageal carcinoma is staged as T1b [27]. The difficulty in diagnosing intramucosal carcinoma is that desmoplasia is not well developed. Features of early invasion include luminal necrosis with apoptotic debris, an architecture with back-to-back glands, the presence of nucleoli in atypical glands, and glands that grow parallel to the surface [28,29]. Intramucosal carcinoma is illustrated in Figs. 17–18. After a diagnosis of intramucosal carcinoma is given on mucosal biopsies, the current standard of care is mucosal resection and ablation, discussed below. Intramucosal carcinoma has an overall favorable prognosis following endoscopic treatment [30]. Selected lesions
3. Basal “crypt” dysplasia This is a controversial diagnosis. This alteration was described in 2006 using the above terminology (the glands in the stomach are pits and not crypts, but since the mucosa is intestinalized the term “crypts” was applied) [25,26]. This pattern makes sense in that biology is a continuum such that we would expect that some cases of dysplasia would be sampled before the findings have reached the surface. In some cases, the nuclear alterations are so striking that these cases can be diagnosed as “high grade dysplasia, basal pattern” without concern as to whether the patient receives mucosal ablation. However, it is less clear which lesions are low grade dysplasia, basal pattern. The authors suggested “lumping” basal pattern dysplasia in the low grade category but at the time of the 2006 publication, universal ablation for low grade dysplasia had yet to be endorsed, so the stakes were smaller than those today. For this reason, some observers prefer to use the indefinite category for such lesions (Figs. 15–16). 3.1. Intramucosal adenocarcinoma and adenocarcinoma In theory intramucosal carcinoma (invasion into the lamina propria only) cannot be diagnosed on mucosal biopsy samples because the
Fig. 16. Basal pattern dysplasia (basal crypt dysplasia). This is a TP53 stain from the area shown in Fig. 15.
Fig. 18. Intramucosal carcinoma. Note that the glands have small protrusions and that there is luminal necrosis. 80
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Table 2 American College of Gastroenterology (ACG) management recommendations. Adapted from Shaheen et al. [3] Grade of dysplasia
Further evaluation
Follow-up interval, ACG/AGA
Likelihood of progression (estimated)
None
None
3–5 y
Indefinite
12 months
Low-grade
Repeat after optimization of acid suppression in 3–6 months. If another indefinite, follow up Expert confirmation followed by ablation
0.2–0.5%/year initially, later up to 9% Unclear
High-grade or intramucosal carcinoma
Expert confirmation. EMR performed for areas of mucosal irregularity coupled with RFA.
Every 6 months during the first year, annually thereafter Every 3 months for first year, every 6 months for second year, then annually thereafter
0.7%/year 7%/year (HGD)
EMR, endoscopic mucosal resection. RFA, radiofrequency ablation.
Fig. 20. Endoscopic mucosal resection specimen. There is an adenocarcinoma invading into the space between the original and duplicated muscularis mucosae that is completely excised. A black dashed line indicates the demarcation between the mucosa and submucosa and a submucosal gland is indicated. Because the sample has curled, the lateral (mucosal) margins are easily mistaken for deep (submucosal) margins. The lateral (mucosal) margins consist of the curled epithelium, lamina propria beneath it, and muscularis mucosae and are indicated by the light blue lines. The deep (submucosal) margin is marked with a red line. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 19. Surgical resection specimen. This is an area of Barrett mucosa in a resection from a patient with adenocarcinoma. There is muscularis propria at the bottom of the image. The submucosa is easy to identify because submucosal glands are present. Above them are both the original and a thick duplicated muscularis mucosae. The layer between the original and duplicated muscularis mucosae is part of the lamina propria and carcinomas invading this layer are staged as T1a.
that invade the very superficial submucosa can also be managed with endoscopic treatments [31,32]. The 2016 American College of Gastroenterologists management guidelines are summarized in Table 2. Mucosal ablation is suggested for any dysplasia including low grade dysplasia, so endoscopy societies recommend expert pathology review before treatment. However, criteria for being an expert are not well established [33]and interobserver variability can be an issue [17] such that it is always reasonable to review cases with colleagues and prudent to have peer-review before performing mucosal ablation. Flat/invisible dysplasia is ablated by radiofrequency ablation whereas dysplasia that forms a visible lesion is resected endoscopically either with endoscopic mucosal resection or, less frequently, endoscopic submucosal dissection. The latter technique provides a larger en bloc sample, usually with negative lateral margins, but the former is easily performed without extensive training and several adjoining areas can be rapidly removed piecemeal [34]. Radiofrequency ablation (RFA [35]) is the preferred ablation method for flat mucosa, but there are other modalities, including cryotherapy [36].
into formalin and they curl. There are three pitfalls that are often encountered in EMRs. Once they are understood, most EMRs are relatively simple to assess. The first is that the surface is often damaged. To perform an EMR, frequently the surface mucosa is sucked into a plastic cap after injecting the submucosa to create a polyp-like lesion that is then removed by polpectomy. The surface is thereby rubbed and sometimes partly denuded. The second artifact is that the sample curls resulting in a misleading appearance that the lateral (mucosal) margin is a deep margin. The third issue is awareness that the muscularis mucosae in Barrett esophagus becomes disorganized and duplicated or even triplicated in response to cycles of damage and repair [37-40]. Based on this, lamina propria can be mistaken for submucosa with resultant over staging of the lesion as T1b instead of T1a. Importantly, the space between the duplicated and original muscularis mucosae is part of the lamina propria. These points are illustrated in Figs. 19–20. When tumors invade the submucosa, providing a depth of invasion as measured from the base of the muscularis mucosae can help colleagues decide whether an esophagectomy should be considered. A depth of more than half a mm (500 μm) often prompts consideration of esophagectomy [32]. The presence of a poorly differentiated component or vascular space invasion also prompts such discussion. Such
3.2. Endoscopic mucosal resection specimens Ideally, endoscopic mucosal resection (EMR) samples should be pinned to a corkboard or wax board, but they are often too small to pin without altering them. Because of this, EMR samples are often dropped 81
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
decisions are best made with a clinical team and in light of known comorbidities in the patient in question. Herein we have summarized some features of Barrett mucosa and patterns of dysplasia and pointed out our method of using features of cell polarity to improve diagnostic consistency in Barrett samples.
[21]
References
[22]
[20]
[1] Fitzgerald RC, di Pietro M, Ragunath K, Ang Y, Kang J-Y, Watson P, et al. British Society of Gastroenterology guidelines on the diagnosis and management of Barrett's oesophagus. Gut 2014;63:7–42. [2] American Gastroenterological A, Spechler SJ, Sharma P, Souza RF, Inadomi JM, Shaheen NJ. American Gastroenterological Association medical position statement on the management of Barrett's esophagus. Gastroenterology 2011;140:1084–91. [3] Shaheen NJ, Falk GW, Iyer PG, Gerson LB, American College of G. ACG clinical guideline: diagnosis and management of Barrett's esophagus. Am J Gastroenterol 2016;111:30–50. [quiz 1]. [4] Aida J, Vieth M, Shepherd NA, Ell C, May A, Neuhaus H, et al. Is carcinoma in columnar-lined esophagus always located adjacent to intestinal metaplasia?: a histopathologic assessment. Am J Surg Pathol 2015;39:188–96. [5] Chandrasoma P, Wijetunge S, DeMeester S, Ma Y, Hagen J, Zamis L, et al. Columnarlined esophagus without intestinal metaplasia has no proven risk of adenocarcinoma. Am J Surg Pathol 2012;36:1–7. [6] Smith J, Garcia A, Zhang R, Demeester S, Vallone J, Chandrasoma P. Intestinal metaplasia is present in most if not all patients who have undergone endoscopic mucosal resection for esophageal adenocarcinoma. Am J Surg Pathol 2016;40:537–43. [7] Salimian KJ, Waters KM, Eze O, Pezhouh MK, Tarabishy Y, Shin EJ, Canto MI, Voltaggio L, Montgomery EA. Definition of Barrett esophagus in the United States: support for retention of a requirement for goblet cells. Am J Surg Pathol 2018;42(2):264–8. [PubMed PMID: 29016405]. [8] Bennett C, Moayyedi P, Corley DA, De Caestecker J, Falck-Ytter Y, Falk G, et al. BOB CAT: a large-scale review and delphi consensus for management of Barrett's esophagus with no dysplasia, indefinite for, or low-grade dysplasia. Am J Gastroenterol 2015;110:662–82. [quiz 83]. [9] Sharma P, Dent J, Armstrong D, Bergman JJ, Gossner L, Hoshihara Y, et al. The development and validation of an endoscopic grading system for Barrett's esophagus: the Prague C & M criteria. Gastroenterology 2006;131:1392–9. [10] Panarelli NC, Yantiss RK. Do ancillary studies aid detection and classification of Barrett esophagus? Am J Surg Pathol 2016;40:e83–93. [11] Ormsby AH, Goldblum JR, Rice TW, Richter JE, Falk GW, Vaezi MF, et al. Cytokeratin subsets can reliably distinguish Barrett's esophagus from intestinal metaplasia of the stomach. Hum Pathol 1999;30:288–94. [12] Ormsby AH, Vaezi MF, Richter JE, Goldblum JR, Rice TW, Falk GW, et al. Cytokeratin immunoreactivity patterns in the diagnosis of short-segment Barrett's esophagus. Gastroenterology 2000;119:683–90. [13] Phillips RW, Frierson Jr. HF, Moskaluk CA. Cdx2 as a marker of epithelial intestinal differentiation in the esophagus. Am J Surg Pathol 2003;27:1442–7. [14] Hahn HP, Blount PL, Ayub K, Das KM, Souza R, Spechler S, et al. Intestinal differentiation in metaplastic, nongoblet columnar epithelium in the esophagus. Am J Surg Pathol 2009;33:1006–15. [15] Chu PG, Jiang Z, Weiss LM. Hepatocyte antigen as a marker of intestinal metaplasia. Am J Surg Pathol 2003;27:952–9. [16] Reid BJ, Haggitt RC, Rubin CE, Roth G, Surawicz CM, Van Belle G, et al. Observer variation in the diagnosis of dysplasia in Barrett's esophagus. Hum Pathol 1988;19:166–78. [17] Montgomery E, Bronner MP, Goldblum JR, Greenson JK, Haber MM, Hart J, et al. Reproducibility of the diagnosis of dysplasia in Barrett esophagus: a reaffirmation. Hum Pathol 2001;32:368–78. [18] Waters K, Salimian K, Voltaggio L, Montgomery EA. Refined criteria for separating low-grade dysplasia and nondysplastic Barrett esophagus reduce equivocal diagnosis and improve prediction of patient outcome: a 10-year review. Am J Surg Pathol 2018. https://doi.org/10.1097/PAS.0000000000001162. PubMed PMID: 30234520 [Epub ahead of print]. [19] Riddell RH, Goldman H, Ransohoff DF, Appelman HD, Fenoglio CM, Haggitt RC,
[23] [24]
[25]
[26]
[27] [28] [29]
[30]
[31]
[32]
[33] [34]
[35]
[36]
[37]
[38]
[39]
[40]
82
et al. Dysplasia in inflammatory bowel disease: standardized classification with provisional clinical applications. Hum Pathol 1983;14:931–68. Curvers WL, ten Kate FJ, Krishnadath KK, Visser M, Elzer B, Baak LC, et al. Lowgrade dysplasia in Barrett's esophagus: overdiagnosed and underestimated. Am J Gastroenterol 2010;105:1523–30. Vieth M, Montgomery EA, Riddell RH. Observations of different patterns of dysplasia in Barrett's esophagus - a first step to harmonize grading. Cesk Patol 2016;52:154–63. Rucker-Schmidt RL, Sanchez CA, Blount PL, Ayub K, Li X, Rabinovitch PS, et al. Nonadenomatous dysplasia in Barrett esophagus: a clinical, pathologic, and DNA content flow cytometric study. Am J Surg Pathol 2009;33:886–93. Mahajan D, Bennett AE, Liu X, Bena J, Bronner MP. Grading of gastric foveolar-type dysplasia in Barrett's esophagus. Mod Pathol 2010;23:1–11. Patil DT, Bennett AE, Mahajan D, Bronner MP. Distinguishing Barrett gastric foveolar dysplasia from reactive cardiac mucosa in gastroesophageal reflux disease. Hum Pathol 2013;44:1146–53. Coco DP, Goldblum JR, Hornick JL, Lauwers GY, Montgomery E, Srivastava A, et al. Interobserver variability in the diagnosis of crypt dysplasia in Barrett esophagus. Am J Surg Pathol 2011;35:45–54. Lomo LC, Blount PL, Sanchez CA, Li X, Galipeau PC, Cowan DS, et al. Crypt dysplasia with surface maturation: a clinical, pathologic, and molecular study of a Barrett's esophagus cohort. Am J Surg Pathol 2006;30:423–35. Amin M, Edge S, Greene F, Schilsky R, Gaspar L, Washington M, et al. AJCC cancer staging manual. 8th ed. Springer International Publishing AG Switzerland; 2017. Voltaggio L, Montgomery EA. Diagnosis and management of Barrett-related neoplasia in the modern era. Surg Pathol Clin 2017;10:781–800. Zhu W, Appelman HD, Greenson JK, Ramsburgh SR, Orringer MB, Chang AC, et al. A histologically defined subset of high-grade dysplasia in Barrett mucosa is predictive of associated carcinoma. Am J Clin Pathol 2009;132:94–100. Pech O, May A, Manner H, Behrens A, Pohl J, Weferling M, et al. Long-term efficacy and safety of endoscopic resection for patients with mucosal adenocarcinoma of the esophagus. Gastroenterology 2014;146:652–60. [e1]. Greene CL, Worrell SG, Attwood SE, Chandrasoma P, Chang K, Demeester TR, et al. Emerging concepts for the endoscopic management of superficial esophageal adenocarcinoma. J Gastrointest Surg 2016;20:851–60. Manner H, Pech O, Heldmann Y, May A, Pauthner M, Lorenz D, et al. The frequency of lymph node metastasis in early-stage adenocarcinoma of the esophagus with incipient submucosal invasion (pT1b sm1) depending on histological risk patterns. Surg Endosc 2015;29:1888–96. van der Wel MJ, Jansen M, Vieth M, Meijer SL. What makes an expert Barrett's histopathologist? Adv Exp Med Biol 2016;908:137–59. Terheggen G, Horn EM, Vieth M, Gabbert H, Enderle M, Neugebauer A, Schumacher B, Neuhaus H. A randomised trial of endoscopic submucosal dissection versus endoscopic mucosal resection for early Barrett's neoplasia. Gut 2017;66(5):783–93. [Epub 2016 Jan 22. PubMed PMID: 26801885; PubMed Central PMCID: PMC5531224]. Shaheen NJ, Sharma P, Overholt BF, Wolfsen HC, Sampliner RE, Wang KK, et al. Radiofrequency ablation in Barrett's esophagus with dysplasia. N Engl J Med 2009;360:2277–88. Canto MI, Shin EJ, Khashab MA, Molena D, Okolo P, Montgomery E, et al. Safety and efficacy of carbon dioxide cryotherapy for treatment of neoplastic Barrett's esophagus. Endoscopy 2015;47:591. Abraham SC, Krasinskas AM, Correa AM, Hofstetter WL, Ajani JA, Swisher SG, et al. Duplication of the muscularis mucosae in Barrett esophagus: an underrecognized feature and its implication for staging of adenocarcinoma. Am J Surg Pathol 2007;31:1719–25. Estrella JS, Hofstetter WL, Correa AM, Swisher SG, Ajani JA, Lee JH, et al. Duplicated muscularis mucosae invasion has similar risk of lymph node metastasis and recurrence-free survival as intramucosal esophageal adenocarcinoma. Am J Surg Pathol 2011;35:1045–53. Kaneshiro DK, Post JC, Rybicki L, Rice TW, Goldblum JR. Clinical significance of the duplicated muscularis mucosae in Barrett esophagus-related superficial adenocarcinoma. Am J Surg Pathol 2011;35:697–700. Lewis JT, Wang KK, Abraham SC. Muscularis mucosae duplication and the musculofibrous anomaly in endoscopic mucosal resections for Barrett esophagus: implications for staging of adenocarcinoma. Am J Surg Pathol 2008;32:566–71.
Contents lists available at ScienceDirect
Annals of Diagnostic Pathology journal homepage: www.elsevier.com/locate/anndiagpath
Review Article
Some observations on Barrett esophagus and associated dysplasia a,⁎
b
c
T
a
Elizabeth Montgomery , Christina A. Arnold , Dora Lam-Himlin , Kevan Salimian , Kevin Watersd a
Department of Pathology, Johns Hopkins University, United States of America Department of Pathology, Ohio State University, United States of America Department of Pathology, Mayo Clinic Scottsdale, United States of America d Department of Pathology, Cedars Sinai Health System, United States of America b c
A R T I C LE I N FO
A B S T R A C T
Keywords: Barrett esophagus Dysplasia Endoscopic mucosal resections Duplicated muscularis mucosae
Biopsy samples from esophageal columnar metaplasia and dysplasia are commonly encountered in Western pathology practice and knowing a few pitfalls can save both pathologists and patients a great deal of anxiety. Herein we discuss criteria for Barrett esophagus, evaluation of dysplasia, and some pitfalls in reviewing endoscopic mucosal resections. Also included is a summary of suggested follow-up for patients with Barrett esophagus.
Barrett esophagus is not defined uniformly worldwide. In the United Kingdom and Japan, the definition differs from that in the United States (US). The British [1], American Gastroenterological Association (AGA) [2], and American College of Gastroenterologists (ACG) [3] criteria appear in Table 1. The key difference between the US and British guidelines is that goblet cells are part of the US definition whereas they are not required in the British definition. However, the 2016 definition from the American College of Gastroenterologists - columnar epithelium with goblet cells extending ≥1 cm above the top of the gastric folds [3] - makes the life of the pathologist challenging. Whereas we frequently have a good idea about the length of a segment of columnar epithelium in question, in other instances, the only information that we receive for a sample is that it is labeled “esophagus”. Of course, if we receive a sample labeled “esophagus, 40 cm” and there is intestinal metaplasia and we have a second sample that is labeled “esophagus, 34 cm” and there is intestinal metaplasia, it is clear that the affected segment of lesion measures at least 1 cm. In fact, the gastroenterology colleagues who prepared the recommendations even went so far as to ask that our endoscopy colleagues refrain from taking biopsies of the gastroesophageal junction in the absence of visible alterations. However, there seems to be very little compliance with the latter suggestion. The American College of Gastroenterology applied the term “specialized intestinal metaplasia of the esophagogastric junction” for lesions with goblet cells that do not meet the 1 cm length requirement [3]. To address the length issue, prepared notes can be useful for situations for which 1) intestinal metaplasia is present without
⁎
knowledge of the segment length or 2) samples labeled “gastroesophageal junction”/GEJ with intestinal metaplasia. 1. Sample notes: Barrett mucosa Situation 1. Barrett mucosa, negative for dysplasia. See note. Note: The above diagnosis of Barrett esophagus is made due to presence of goblet cells (intestinal metaplasia) with the assumption that the biopsies were obtained from columnar mucosa in the distal esophagus and the mucosal irregularity extends at least 1 cm above the top of the gastric folds as per 2016 American College of Gastroenterology (ACG) guidelines. Reference: Shaheen NJ, Falk GW, Iyer PG, Gerson LB; American College of Gastroenterology. ACG Clinical Guideline: Diagnosis and Management of Barrett's Esophagus. Am J Gastroenterol. 2016 Jan;111(1):30–50. Situation 2. Cardiac mucosa with intestinal metaplasia. See note. Note: This biopsy shows gastric-type mucosa with scattered goblet cells. The diagnosis in this case depends on the location of this biopsy. If this biopsy was taken from the tubular esophagus and the mucosal irregularity extends at least 1 cm above the top of the gastric folds, it shows Barrett mucosa of the distinctive type. If this biopsy was taken from the gastric cardia, it shows intestinal metaplasia of the gastric cardia. Reference: Shaheen NJ, Falk GW, Iyer PG, Gerson LB; American College of Gastroenterology. ACG Clinical Guideline: Diagnosis and
Corresponding author at: 410 North Broadway, Weinberg Building Room 2242, Baltimore, MD 21231, United States of America. E-mail address: [email protected] (E. Montgomery).
https://doi.org/10.1016/j.anndiagpath.2018.09.013 Received 23 September 2018; Accepted 26 September 2018 1092-9134/ © 2018 Elsevier Inc. All rights reserved.
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Table 1 Definition of Barrett esophagus. Society
Definition
British (and Japanese) definition of Barrett mucosa 2014 American Gastroenterological Association definition of Barrett mucosa 2011 American College of Gastroenterologists' definition of Barrett mucosa 2016
- Columnar epithelium with or without goblet cells extending ≥1 cm above the gastric folds [1] - Columnar epithelium in the esophagus that contains goblet cells – no length requirement [2] - Columnar epithelium with goblet cells extending ≥1 cm above the top of the gastric folds [3]
[11,12]. These authors studied long-segment Barrett esophagus cases (> 3 cm) and noted superficial and deep CK7 immunoreactivity in the intestinalized mucosa, with only superficial CK20 staining in the intestinalized zones. In contrast, distal gastric intestinal metaplasia showed patchy, superficial, and deep CK20 staining in areas of incomplete intestinal metaplasia; strong, superficial, and deep CK20 staining in areas of complete intestinal metaplasia; and patchy or absent CK7 staining in either type of gastric intestinal metaplasia. Other studies have not confirmed these findings and CK7/20 immunolabeling has fallen out of favor. Other immunostains that have been studied include mucin core (MUC) polypeptides, which seem to be of little practical value. However, the key markers are MUC5 (gastric foveolar mucin), MUC6 (cardiac glands, antral glands, Brunner glands), and MUC2 (goblet cells). CDX2 staining has also been used to label areas of intestinal metaplasia [13] and some have noted that cases lacking goblet cells express these markers and believe that this supports the need to eliminate the requirement for goblet cells [14]. Hepatocyte antigen (Hepar-1, Carbamoyl Phosphate Synthetase 1) also marks intestinalization in the absence of goblet cells [15]. However, in daily practice it is more practical to simply search for goblet cells. None of the immunostains offers added value over H&E stains in detecting goblet cells [10].
Management of Barrett's Esophagus. Am J Gastroenterol. 2016 Jan;111(1):30–50. There are some observers in the US who have suggested eliminating the requirement for goblet cells in diagnosing Barrett mucosa since some esophageal adenocarcinomas arise in the absence of intestinal metaplasia. As an example, in one study, the authors reviewed endoscopic mucosal resection samples from a cohort of German patients and found adjoining intestinal metaplasia in association with less than half with early cancers [4]. However, these authors made no attempt to learn if the patients had separate samples with intestinal metaplasia. In two subsequent studies from the US West Coast, high grade columnar epithelial dysplasia and carcinomas were essentially always accompanied by intestinal metaplasia [5,6]. Similar results were found in an East Coast study [7] such that we would endorse retaining the requirement for goblet cells, a view not shared by all [8]. Indeed, there are some examples of esophageal adenocarcinomas that are unassociated with intestinal metaplasia but these are not numerous in our Western population. Eliminating the requirement for goblet cells would even further reduce an already unfavorable cost effectiveness of surveillance for esophageal adenocarcinoma. Endoscopically, Barrett mucosa consists of velvety “salmon colored” epithelium that can as extend as “tongues” above the gastric folds (Fig. 1). When an area of Barrett mucosa is surrounded by squamous mucosa, the appearance is referred to as an “island”. Our endoscopy colleagues use the Prague system to describe the extent of Barrett mucosa. In this system, the distance of the circumferential length of Barrett mucosa is recorded (“C”) and the maximum length is recorded as well (“M”) [9]. This method allows standardization of endoscopy reports and when these data are provided to pathologists, they afford some confidence in our diagnoses. In general, no special stains are needed to confirm the presence of goblet cells in esophageal biopsies. This topic was comprehensively reviewed by Panarelli and Yantiss, who concluded that neither histochemical nor immunohistochemical stains add value over H&E stains since they produce false positives [10]. In the past, the concept of using CK7/CK20 stains to separate esophageal intestinal metaplasia from gastric cardiac intestinal metaplasia was introduced by Ormsby et al.
2. Grading Barrett dysplasia The categories that are used to interpret biopsies [16]: Negative for dysplasia Indefinite for dysplasia Low grade dysplasia High grade dysplasia Adenocarcinoma Assessing Barrett biopsies is usually straightforward since most cases are nondysplastic but it is well known that observer variation can be an issue [17]. We have tightened our criteria in the last few years using a novel but very simple method to assess cell polarity (the relationship of cells one to another). We have also suggested that the combined number of cases diagnosed as indefinite for dysplasia, low grade dysplasia, and high grade dysplasia should not exceed 10% [18]. Of course, those clinics specializing in dysplasia would be expected to have a higher percentage of dysplasia cases. In evaluating Barrett mucosa, essentially the sample should be assessed for surface maturation and glandular crowding, its cytologic features, and whether inflammation is an obscuring factor before making a diagnosis. 2.1. Barrett esophagus, negative for dysplasia Nondysplastic Barrett mucosa should show surface maturation, which can be a challenge to confirm in suboptimally embedded samples. Minor nuclear alterations in the bases of the metaplastic pits are acceptable. However, noting the polarity of the epithelial cells and how they are arranged with respect to one another makes assessment of Barrett mucosa relatively easy. We have assessed Barrett mucosa easily in most cases by paying attention to “the four lines” [18]. Finding “the four lines” indicates preserved polarity of epithelial cells in both gastric
Fig. 1. This is an endocropic image from a patient with Barrett esophagus. The gastric folds are seen at the left of the image and a tongue of metaplastic epithelium is present. Note that the squamous epithelium at the right has a greyish pearly appearance. 76
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 2. Barrett mucosa with no dysplasia. Although scattered goblet cells are present, the area with the arrows indicating “the lines” lacks them and has an appearance like that of gastric cardiac surface epithelium. The top line is formed by the apical neutral mucin cap and the second by the base of that cap. The third line consists of the eosinophilic cytoplasm and the fourth by the layer of nuclei.
Fig. 4. Barrett esophagus indefinite for dysplasia. Inflammation and suboptimal sectioning are both factors. The surface nuclei are hyperchromatic and surface “lines” are somewhat obscured, particularly at the upper left of the image but the nuclei are not particularly enlarged.
reduce any inflammation followed by repeat sampling [3]. Of course no matter which images we show to illustrate our conception of cases that we would regard as indefinite for dysplasia, some colleagues will dismiss the illustrated changes as reactive and others will be concerned that dysplasia has been overlooked. That is the entire point of this category. It merits follow-up but the patient should not receive definitive treatment (mucosal ablation therapy) until the findings are clarified. The indefinite category tends to reflect lesions that are inflamed, situated in the bases of the pits with surface maturation, or for which artifact (diathermy) impede interpretation (Figs. 4–5). This indefinite category should be used in only a small percentage of cases (up to 3–5% but this is high) In fact, some colleagues essentially never use this category.
2.3. Low grade dysplasia Fig. 3. Barrett mucosa with no dysplasia. The lines described in Fig. 2 are indicated. In this area, the lines are interrupted by goblet cells.
Low grade dysplasia should be unequivocally neoplastic (adenomalike) and involve the surface epithelium. It is important not to overdiagnose low grade dysplasia as current guidelines endorse mucosal
cardiac mucosa and in Barrett mucosa (with goblet cells). Cardiac mucosa and Barrett mucosa are annotated in Figs. 2–3 to demonstrate the four lines at high magnification, but they can be identified at 4×. The top line comprises the apical mucin cap of neutral mucin typical of Barrett mucosa (with incomplete intestinal metaplasia) and cardiac mucosa. The second is composed of the bases of the mucin caps. The third reflects the aligned cytoplasm and the last (fourth) line is formed by the row of nuclei. Attention to these patterns makes it easy to dismiss many cases as reactive. 2.2. Barrett esophagus indefinite for dysplasia The concept of uncertainty in diagnosis was introduced in a 1983 study of epithelial changes in inflammatory bowel disease and was a radical change [19] since it was incomprehensible to our clinical colleagues that sometimes we are unable to discern whether epithelial changes are dysplastic/neoplastic or not. The term was later applied to columnar epithelial changes in Barrett mucosa [16,17] and, until recently, gastroenterology societies did not attempt to address it in guidelines. This category is presently regarded as a holding diagnosis until the findings in the patient are clarified by a course of treatment to
Fig. 5. Barrett esophagus indefinite for dysplasia. This is a higher magnification of the lesion seen in Fig. 4. 77
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 6. Low grade dysplasia. The surface nuclei hyperchromatic (adenoma-like) and there is an abrupt transition (arrow) between the zone in question and the remaining tissue.
Fig. 8. High grade dysplasia. Note the jumbled appearance of the nuclei at the surface. The nuclei are quite hyperchromatic and there is plenty of nuclear stratification.
ablation for low grade dysplasia and mucosal ablation confers a risk for stricture formation. In prospectively evaluated patients presenting to a clinic, the incidence of low grade dysplasia should be about 2–3% but not > 5% [20]. The nuclei in low grade dysplasia are larger than those of nondysplastic Barrett mucosa and generally there is little inflammation in samples confidently diagnosed with low grade dysplasia. A very helpful clue is an abrupt transition between the dysplastic zone and adjoining zones that are clearly not dysplastic (Figs. 6–7). The surface lines that characterize nondysplastic Barrett mucosa are effaced. Some examples of low-grade dysplasia can resemble colorectal tubular adenomas. Low grade dysplasia demonstrates loss of the four lines (an overall indication of altered cell architecture and cell polarity) but maintained nuclear polarity such that the long axes of the nuclei should remain more or less perpendicular to the basement membrane. We would note that there is some subjectivity in separating low grade dysplasia and high grade dysplasia but this is less important than it was in the past; all dysplasia is currently managed by endoscopic ablation [1,3]. Staining for TP53 is sometimes useful in confirming an interpretation of low grade (as well as of high grade) dysplasia - a few darkly stained nuclei on the surface can be identified and the basal glands label.
Most dysplasia cases show intestinal differentiation in that the epithelium is stratified and contains goblet cells - similar to the appearance of colorectal adenomas, but not all do [21]. 2.4. High grade dysplasia High grade dysplasia is usually easy to recognize. The area appears hyperchromatic at low magnification, standing out from any nondysplastic mucosa in the sample. The alterations are usually detected in the surface epithelium. Prominent nucleoli are not typical of high grade dysplasia. Usually there is ample lamina propria between the glands. Many nuclei both in the pits and at the surface have lost their relationship to the basement membrane and have rounded up and are arranged in a jumbled configuration (Fig. 8). Most cases of high grade dysplasia show some nuclear elongation and stratification similar to that in colorectal adenomas. However, some cases show an unusual pattern of small tubules, each lined by a monolayer of hyperchromatic nuclei (Fig. 9). Since the monolayer lacks nuclear stratification, it has been referred to as a “nonadenomatous” form of high grade dysplasia by some [22] whereas others have considered this pattern as evidence for gastric differentiation in high grade
Fig. 9. High grade dysplasia. The indicated glands (arrows) that have a monolayer of round hyperchromatic nuclei are dysplastic (high grade). Some examples of high dysplasia show only this type of glands.
Fig. 7. Low grade dysplasia. This is a higher magnification of the process seen in Fig. 6 showing the abrupt transition (arrow). 78
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 10. High grade dysplasia. TP53 immunolabeling labels the nuclei intensely.
Fig. 12. High grade dysplasia. This is a TP53 stain from the lesion seen in Fig. 11. There is light labeling in the nuclei of the normal cardiac glands (wild type pattern) and some of the lymphoid cells but complete absence of labeling in the dysplastic glands, suggesting biallelic inactivation of the TP53 gene.
dysplasia [23,24]. Immunolabeling for TP53 can be helpful to confirm an impression of high grade dysplasia when there is doubt. If the histologic features are already classic, performing immunolabeling adds no value, but about 85–90% of the time, if immunolabeling is added, strong labeling confirms the diagnosis (Fig. 10). TP53 is a tumor suppressor protein normally active during cell division and has a short half-life. Since TP53 is a normal molecule, some labeling is to be expected in the proliferative compartment of the mucosa. Therefore there is nuclear labeling in the basal layer of the squamous epithelium and in the pits of the stomach, the proliferative compartments of these tissues. When the TP53 gene is mutated, this results in a TP53 protein with a long half-life, so the protein accumulates in the nuclei of the cells and can be detected by immunolabeling. There is also a subset of high grade dysplasia cases in which biallelic loss of the TP53 gene results in complete absence of the protein, termed the “null pattern”. This can be exploited for diagnosis just as well as the strongly reactive pattern (Figs. 11–12). Finally, there are cases of both low grade and high grade that display either gastric foveolar differentiation or differentiation similar to that seen in pyloric gland adenomas This type of dysplasia can arise in a background of intestinal metaplasia but it still has gastric type differentiation [21]. The nuclei in this pattern of dysplasia are paler than those seen in conventional type dysplasia and this form of dysplasia is
Fig. 13. Foveolar pattern low grade dysplasia. Although there are a few goblet cells, the low grade dysplasia seen at the surface is characterized by enlarged nuclei and loss of “the lines”.
Fig. 14. Foveolar pattern low grade dysplasia. This is a higher magnification of the lesion shown in Fig. 13.
Fig. 11. High grade dysplasia. This image allows comparison between the dysplastic nuclei and the nuclei of a few cardiac glands. 79
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Fig. 15. Basal pattern dysplasia (basal crypt dysplasia). Note the pocket at the lower left, which jumps out because the nuclei are strikingly hyperchromatic. There is not unequivocal surface involvement. Fig. 17. Intramucosal carcinoma. There are glands with luminal necrosis and the one in the upper center of the lesion grows parallel rather than perpendicular to the surface.
diagnosed by attention to nuclear enlargement (low grade dysplasia) (Figs. 13–14); and mutilayering and loss of nuclear polarity in high grade dysplasia.
tissue beneath the mucosa is not present for evaluation, but a good estimate can be made on mucosal biopsies alone! Because of this, we diagnose intramucosal carcinoma (invasion into the lamina propria) on mucosal samples because we know that this interpretation will prompt an endoscopic mucosal reaction we can then better assess for deeper invasion. However, some institutions diagnose “adenocarcinoma” with a note that the patient should be evaluated clinically to determine whether the lesion can be managed endoscopically. Intramucosal carcinoma denotes invasion into the mucosa and thus is staged as T1a [27], which differs from the colorectal staging scheme for which lamina propria invasion is staged as Tis because there is minimal lymphatic access in the lamina propria of the colon. Invasion into the submucosa by esophageal carcinoma is staged as T1b [27]. The difficulty in diagnosing intramucosal carcinoma is that desmoplasia is not well developed. Features of early invasion include luminal necrosis with apoptotic debris, an architecture with back-to-back glands, the presence of nucleoli in atypical glands, and glands that grow parallel to the surface [28,29]. Intramucosal carcinoma is illustrated in Figs. 17–18. After a diagnosis of intramucosal carcinoma is given on mucosal biopsies, the current standard of care is mucosal resection and ablation, discussed below. Intramucosal carcinoma has an overall favorable prognosis following endoscopic treatment [30]. Selected lesions
3. Basal “crypt” dysplasia This is a controversial diagnosis. This alteration was described in 2006 using the above terminology (the glands in the stomach are pits and not crypts, but since the mucosa is intestinalized the term “crypts” was applied) [25,26]. This pattern makes sense in that biology is a continuum such that we would expect that some cases of dysplasia would be sampled before the findings have reached the surface. In some cases, the nuclear alterations are so striking that these cases can be diagnosed as “high grade dysplasia, basal pattern” without concern as to whether the patient receives mucosal ablation. However, it is less clear which lesions are low grade dysplasia, basal pattern. The authors suggested “lumping” basal pattern dysplasia in the low grade category but at the time of the 2006 publication, universal ablation for low grade dysplasia had yet to be endorsed, so the stakes were smaller than those today. For this reason, some observers prefer to use the indefinite category for such lesions (Figs. 15–16). 3.1. Intramucosal adenocarcinoma and adenocarcinoma In theory intramucosal carcinoma (invasion into the lamina propria only) cannot be diagnosed on mucosal biopsy samples because the
Fig. 16. Basal pattern dysplasia (basal crypt dysplasia). This is a TP53 stain from the area shown in Fig. 15.
Fig. 18. Intramucosal carcinoma. Note that the glands have small protrusions and that there is luminal necrosis. 80
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
Table 2 American College of Gastroenterology (ACG) management recommendations. Adapted from Shaheen et al. [3] Grade of dysplasia
Further evaluation
Follow-up interval, ACG/AGA
Likelihood of progression (estimated)
None
None
3–5 y
Indefinite
12 months
Low-grade
Repeat after optimization of acid suppression in 3–6 months. If another indefinite, follow up Expert confirmation followed by ablation
0.2–0.5%/year initially, later up to 9% Unclear
High-grade or intramucosal carcinoma
Expert confirmation. EMR performed for areas of mucosal irregularity coupled with RFA.
Every 6 months during the first year, annually thereafter Every 3 months for first year, every 6 months for second year, then annually thereafter
0.7%/year 7%/year (HGD)
EMR, endoscopic mucosal resection. RFA, radiofrequency ablation.
Fig. 20. Endoscopic mucosal resection specimen. There is an adenocarcinoma invading into the space between the original and duplicated muscularis mucosae that is completely excised. A black dashed line indicates the demarcation between the mucosa and submucosa and a submucosal gland is indicated. Because the sample has curled, the lateral (mucosal) margins are easily mistaken for deep (submucosal) margins. The lateral (mucosal) margins consist of the curled epithelium, lamina propria beneath it, and muscularis mucosae and are indicated by the light blue lines. The deep (submucosal) margin is marked with a red line. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
Fig. 19. Surgical resection specimen. This is an area of Barrett mucosa in a resection from a patient with adenocarcinoma. There is muscularis propria at the bottom of the image. The submucosa is easy to identify because submucosal glands are present. Above them are both the original and a thick duplicated muscularis mucosae. The layer between the original and duplicated muscularis mucosae is part of the lamina propria and carcinomas invading this layer are staged as T1a.
that invade the very superficial submucosa can also be managed with endoscopic treatments [31,32]. The 2016 American College of Gastroenterologists management guidelines are summarized in Table 2. Mucosal ablation is suggested for any dysplasia including low grade dysplasia, so endoscopy societies recommend expert pathology review before treatment. However, criteria for being an expert are not well established [33]and interobserver variability can be an issue [17] such that it is always reasonable to review cases with colleagues and prudent to have peer-review before performing mucosal ablation. Flat/invisible dysplasia is ablated by radiofrequency ablation whereas dysplasia that forms a visible lesion is resected endoscopically either with endoscopic mucosal resection or, less frequently, endoscopic submucosal dissection. The latter technique provides a larger en bloc sample, usually with negative lateral margins, but the former is easily performed without extensive training and several adjoining areas can be rapidly removed piecemeal [34]. Radiofrequency ablation (RFA [35]) is the preferred ablation method for flat mucosa, but there are other modalities, including cryotherapy [36].
into formalin and they curl. There are three pitfalls that are often encountered in EMRs. Once they are understood, most EMRs are relatively simple to assess. The first is that the surface is often damaged. To perform an EMR, frequently the surface mucosa is sucked into a plastic cap after injecting the submucosa to create a polyp-like lesion that is then removed by polpectomy. The surface is thereby rubbed and sometimes partly denuded. The second artifact is that the sample curls resulting in a misleading appearance that the lateral (mucosal) margin is a deep margin. The third issue is awareness that the muscularis mucosae in Barrett esophagus becomes disorganized and duplicated or even triplicated in response to cycles of damage and repair [37-40]. Based on this, lamina propria can be mistaken for submucosa with resultant over staging of the lesion as T1b instead of T1a. Importantly, the space between the duplicated and original muscularis mucosae is part of the lamina propria. These points are illustrated in Figs. 19–20. When tumors invade the submucosa, providing a depth of invasion as measured from the base of the muscularis mucosae can help colleagues decide whether an esophagectomy should be considered. A depth of more than half a mm (500 μm) often prompts consideration of esophagectomy [32]. The presence of a poorly differentiated component or vascular space invasion also prompts such discussion. Such
3.2. Endoscopic mucosal resection specimens Ideally, endoscopic mucosal resection (EMR) samples should be pinned to a corkboard or wax board, but they are often too small to pin without altering them. Because of this, EMR samples are often dropped 81
Annals of Diagnostic Pathology 37 (2018) 75–82
E. Montgomery et al.
decisions are best made with a clinical team and in light of known comorbidities in the patient in question. Herein we have summarized some features of Barrett mucosa and patterns of dysplasia and pointed out our method of using features of cell polarity to improve diagnostic consistency in Barrett samples.
[21]
References
[22]
[20]
[1] Fitzgerald RC, di Pietro M, Ragunath K, Ang Y, Kang J-Y, Watson P, et al. British Society of Gastroenterology guidelines on the diagnosis and management of Barrett's oesophagus. Gut 2014;63:7–42. [2] American Gastroenterological A, Spechler SJ, Sharma P, Souza RF, Inadomi JM, Shaheen NJ. American Gastroenterological Association medical position statement on the management of Barrett's esophagus. Gastroenterology 2011;140:1084–91. [3] Shaheen NJ, Falk GW, Iyer PG, Gerson LB, American College of G. ACG clinical guideline: diagnosis and management of Barrett's esophagus. Am J Gastroenterol 2016;111:30–50. [quiz 1]. [4] Aida J, Vieth M, Shepherd NA, Ell C, May A, Neuhaus H, et al. Is carcinoma in columnar-lined esophagus always located adjacent to intestinal metaplasia?: a histopathologic assessment. Am J Surg Pathol 2015;39:188–96. [5] Chandrasoma P, Wijetunge S, DeMeester S, Ma Y, Hagen J, Zamis L, et al. Columnarlined esophagus without intestinal metaplasia has no proven risk of adenocarcinoma. Am J Surg Pathol 2012;36:1–7. [6] Smith J, Garcia A, Zhang R, Demeester S, Vallone J, Chandrasoma P. Intestinal metaplasia is present in most if not all patients who have undergone endoscopic mucosal resection for esophageal adenocarcinoma. Am J Surg Pathol 2016;40:537–43. [7] Salimian KJ, Waters KM, Eze O, Pezhouh MK, Tarabishy Y, Shin EJ, Canto MI, Voltaggio L, Montgomery EA. Definition of Barrett esophagus in the United States: support for retention of a requirement for goblet cells. Am J Surg Pathol 2018;42(2):264–8. [PubMed PMID: 29016405]. [8] Bennett C, Moayyedi P, Corley DA, De Caestecker J, Falck-Ytter Y, Falk G, et al. BOB CAT: a large-scale review and delphi consensus for management of Barrett's esophagus with no dysplasia, indefinite for, or low-grade dysplasia. Am J Gastroenterol 2015;110:662–82. [quiz 83]. [9] Sharma P, Dent J, Armstrong D, Bergman JJ, Gossner L, Hoshihara Y, et al. The development and validation of an endoscopic grading system for Barrett's esophagus: the Prague C & M criteria. Gastroenterology 2006;131:1392–9. [10] Panarelli NC, Yantiss RK. Do ancillary studies aid detection and classification of Barrett esophagus? Am J Surg Pathol 2016;40:e83–93. [11] Ormsby AH, Goldblum JR, Rice TW, Richter JE, Falk GW, Vaezi MF, et al. Cytokeratin subsets can reliably distinguish Barrett's esophagus from intestinal metaplasia of the stomach. Hum Pathol 1999;30:288–94. [12] Ormsby AH, Vaezi MF, Richter JE, Goldblum JR, Rice TW, Falk GW, et al. Cytokeratin immunoreactivity patterns in the diagnosis of short-segment Barrett's esophagus. Gastroenterology 2000;119:683–90. [13] Phillips RW, Frierson Jr. HF, Moskaluk CA. Cdx2 as a marker of epithelial intestinal differentiation in the esophagus. Am J Surg Pathol 2003;27:1442–7. [14] Hahn HP, Blount PL, Ayub K, Das KM, Souza R, Spechler S, et al. Intestinal differentiation in metaplastic, nongoblet columnar epithelium in the esophagus. Am J Surg Pathol 2009;33:1006–15. [15] Chu PG, Jiang Z, Weiss LM. Hepatocyte antigen as a marker of intestinal metaplasia. Am J Surg Pathol 2003;27:952–9. [16] Reid BJ, Haggitt RC, Rubin CE, Roth G, Surawicz CM, Van Belle G, et al. Observer variation in the diagnosis of dysplasia in Barrett's esophagus. Hum Pathol 1988;19:166–78. [17] Montgomery E, Bronner MP, Goldblum JR, Greenson JK, Haber MM, Hart J, et al. Reproducibility of the diagnosis of dysplasia in Barrett esophagus: a reaffirmation. Hum Pathol 2001;32:368–78. [18] Waters K, Salimian K, Voltaggio L, Montgomery EA. Refined criteria for separating low-grade dysplasia and nondysplastic Barrett esophagus reduce equivocal diagnosis and improve prediction of patient outcome: a 10-year review. Am J Surg Pathol 2018. https://doi.org/10.1097/PAS.0000000000001162. PubMed PMID: 30234520 [Epub ahead of print]. [19] Riddell RH, Goldman H, Ransohoff DF, Appelman HD, Fenoglio CM, Haggitt RC,
[23] [24]
[25]
[26]
[27] [28] [29]
[30]
[31]
[32]
[33] [34]
[35]
[36]
[37]
[38]
[39]
[40]
82
et al. Dysplasia in inflammatory bowel disease: standardized classification with provisional clinical applications. Hum Pathol 1983;14:931–68. Curvers WL, ten Kate FJ, Krishnadath KK, Visser M, Elzer B, Baak LC, et al. Lowgrade dysplasia in Barrett's esophagus: overdiagnosed and underestimated. Am J Gastroenterol 2010;105:1523–30. Vieth M, Montgomery EA, Riddell RH. Observations of different patterns of dysplasia in Barrett's esophagus - a first step to harmonize grading. Cesk Patol 2016;52:154–63. Rucker-Schmidt RL, Sanchez CA, Blount PL, Ayub K, Li X, Rabinovitch PS, et al. Nonadenomatous dysplasia in Barrett esophagus: a clinical, pathologic, and DNA content flow cytometric study. Am J Surg Pathol 2009;33:886–93. Mahajan D, Bennett AE, Liu X, Bena J, Bronner MP. Grading of gastric foveolar-type dysplasia in Barrett's esophagus. Mod Pathol 2010;23:1–11. Patil DT, Bennett AE, Mahajan D, Bronner MP. Distinguishing Barrett gastric foveolar dysplasia from reactive cardiac mucosa in gastroesophageal reflux disease. Hum Pathol 2013;44:1146–53. Coco DP, Goldblum JR, Hornick JL, Lauwers GY, Montgomery E, Srivastava A, et al. Interobserver variability in the diagnosis of crypt dysplasia in Barrett esophagus. Am J Surg Pathol 2011;35:45–54. Lomo LC, Blount PL, Sanchez CA, Li X, Galipeau PC, Cowan DS, et al. Crypt dysplasia with surface maturation: a clinical, pathologic, and molecular study of a Barrett's esophagus cohort. Am J Surg Pathol 2006;30:423–35. Amin M, Edge S, Greene F, Schilsky R, Gaspar L, Washington M, et al. AJCC cancer staging manual. 8th ed. Springer International Publishing AG Switzerland; 2017. Voltaggio L, Montgomery EA. Diagnosis and management of Barrett-related neoplasia in the modern era. Surg Pathol Clin 2017;10:781–800. Zhu W, Appelman HD, Greenson JK, Ramsburgh SR, Orringer MB, Chang AC, et al. A histologically defined subset of high-grade dysplasia in Barrett mucosa is predictive of associated carcinoma. Am J Clin Pathol 2009;132:94–100. Pech O, May A, Manner H, Behrens A, Pohl J, Weferling M, et al. Long-term efficacy and safety of endoscopic resection for patients with mucosal adenocarcinoma of the esophagus. Gastroenterology 2014;146:652–60. [e1]. Greene CL, Worrell SG, Attwood SE, Chandrasoma P, Chang K, Demeester TR, et al. Emerging concepts for the endoscopic management of superficial esophageal adenocarcinoma. J Gastrointest Surg 2016;20:851–60. Manner H, Pech O, Heldmann Y, May A, Pauthner M, Lorenz D, et al. The frequency of lymph node metastasis in early-stage adenocarcinoma of the esophagus with incipient submucosal invasion (pT1b sm1) depending on histological risk patterns. Surg Endosc 2015;29:1888–96. van der Wel MJ, Jansen M, Vieth M, Meijer SL. What makes an expert Barrett's histopathologist? Adv Exp Med Biol 2016;908:137–59. Terheggen G, Horn EM, Vieth M, Gabbert H, Enderle M, Neugebauer A, Schumacher B, Neuhaus H. A randomised trial of endoscopic submucosal dissection versus endoscopic mucosal resection for early Barrett's neoplasia. Gut 2017;66(5):783–93. [Epub 2016 Jan 22. PubMed PMID: 26801885; PubMed Central PMCID: PMC5531224]. Shaheen NJ, Sharma P, Overholt BF, Wolfsen HC, Sampliner RE, Wang KK, et al. Radiofrequency ablation in Barrett's esophagus with dysplasia. N Engl J Med 2009;360:2277–88. Canto MI, Shin EJ, Khashab MA, Molena D, Okolo P, Montgomery E, et al. Safety and efficacy of carbon dioxide cryotherapy for treatment of neoplastic Barrett's esophagus. Endoscopy 2015;47:591. Abraham SC, Krasinskas AM, Correa AM, Hofstetter WL, Ajani JA, Swisher SG, et al. Duplication of the muscularis mucosae in Barrett esophagus: an underrecognized feature and its implication for staging of adenocarcinoma. Am J Surg Pathol 2007;31:1719–25. Estrella JS, Hofstetter WL, Correa AM, Swisher SG, Ajani JA, Lee JH, et al. Duplicated muscularis mucosae invasion has similar risk of lymph node metastasis and recurrence-free survival as intramucosal esophageal adenocarcinoma. Am J Surg Pathol 2011;35:1045–53. Kaneshiro DK, Post JC, Rybicki L, Rice TW, Goldblum JR. Clinical significance of the duplicated muscularis mucosae in Barrett esophagus-related superficial adenocarcinoma. Am J Surg Pathol 2011;35:697–700. Lewis JT, Wang KK, Abraham SC. Muscularis mucosae duplication and the musculofibrous anomaly in endoscopic mucosal resections for Barrett esophagus: implications for staging of adenocarcinoma. Am J Surg Pathol 2008;32:566–71.