Adenocarcinoma in Barrett mucosa treated by endoscopic mucosal resection.
According to guidelines used in the United States, there are both endoscopic and histologic requirements for the diagnosis of Barrett mucosa.1 The endoscopic requirement is the presence of an abnormal columnar-appearing mucosa in the lower esophagus, and the histologic requirement is the presence of goblet cells in the biopsy specimen of the abnormal endoscopic mucosa. Every so often, a patient's endoscopist may give a diagnosis of Barrett mucosa, although the biopsy specimen has the presence of only dysplastic mucosa with no Barrett mucosa. Any nondysplastic epithelium is either squamous, from the distal squamous esophagus, or gastric cardiac-type mucosa. This raises a new issue, namely, do the same guidelines apply to a dysplastic biopsy specimen that an endoscopist associated with Barrett mucosa, although histologically there is no evidence of Barrett mucosa (and therefore the biopsy is basically that of a gastroesophageal junction or a cardiac biopsy)? No one has specifically dealt with this issue. It appears that these patients are managed as if they had Barrett mucosa.
The introduction of new treatment modalities by our clinical colleagues has further changed the clinical approaches to dysplastic epithelium. The new paradigm includes ablation technology, whereby, using a variety of injurious methods, the abnormal mucosa is destroyed. Furthermore, there are types of limited resections, which are referred to as endoscopic mucosal resections (EMRs), in which the abnormal mucosa is removed endoscopically, such that no surgical resections are required unless the lesion cannot be removed endoscopically. (2-4) Endoscopic mucosal resections are usually performed for some visible abnormality. In one such approach, the endoscopist injects a liquid, such as saline, into the submucosa beneath the abnormal mucosa, thereby elevating it and basically turning it into a polyp. The abnormality is then removed by polypectomy by using a cautery loop. Endoscopic mucosal resection traumatizes the tissue. In the resultant specimens, we have to deal with procedural artifacts, including shredding of the surface epithelium, dissociation of deeper epithelium, cautery, and crushing. Overall, the specimen is often a light microscopic "mess." The entire peripheral and deep margins are cauterized. Cautery is a useful marker because it indicates whether the microscopic section includes these margins; however, it destroys any tissue detail at these margins, which may be important for diagnosis, such as dysplasia and carcinoma.
[FIGURE 1 OMITTED]
REPORT OF A CASE
The patient is a middle aged man with chronic heartburn, who was treated with protein pump inhibitors but with inadequate response. An upper endoscopy was performed, with the finding of a Barrett segment of 1 cm. A biopsy specimen of the segment was interpreted as invasive adenocarcinoma, but there was no Barrett mucosa. A repeated upper endoscopy was performed, and this time the endoscopist found a 9 mm nodule at the gastroesophageal junction, presumably the site of the carcinoma in the original biopsy. The carcinoma was still present and there was still no evidence of Barrett mucosa. An endoscopic ultrasonography was performed to see how deep the carcinoma had invaded the tissue, but because the ultrasonographer did not detect the carcinoma, it appeared that the carcinoma had not invaded the tissue deeply. Therefore, an endoscopic mucosal resection of the nodule was performed. The resected specimen was cut into five 2-mm thick pieces, embedded, and sectioned. There was squamous mucosa proximally and cardiac-type columnar mucosa distally. The sections with cardiac-type mucosa did not have goblet cells. There was neoplastic mucosa, including both dysplasias and carcinoma. Was Barrett mucosa originally present that had become dysplastic and could no longer be found? Perhaps. In this case, much of the epithelium on the surface and lining the underlying tubules had full-thickness nuclear stratification and very little cytoplasmic maturation and was fragmented from the EMR procedure. Some of the tubules had complex architecture, and overall looked like HGD (Figure 1). How does one know it is HGD? I use criteria. But exactly what are these criteria? I will now interrupt the case report with a comment that deals with the criteria for dysplasia, invasive carcinoma, and anatomic changes unique to Barrett mucosa.
What are the implications for diagnoses for the 2 grades of dysplasia? A diagnosis of LGD generally leads to follow-up rather than intervention, while a diagnosis of HGD is much more likely to lead to intervention. Therefore, we need to know the minimal criteria for the diagnosis of HGD in Barrett mucosa. We need to be able to tell the difference between the highest end of LGD dysplasia and the lowest end of HGD. My idealized LGD has nuclei with little variation that are confined mainly to the basal half of the cells, have obvious cytoplasmic mucinous maturation, and the overall impression is one of uniformity, that is, the cells all look very much alike. In contrast, my idealized HGD is epithelium that has full-thickness nuclear stratification, pleomorphism, no cytoplasmic maturation, with the overall impression being that this is not a uniform epithelium. What features should be added to this idealized LGD to place it, albeit barely, into the high-grade category, or what features need to be subtracted from this idealized HGD to place it into the very top of the low-grade category? We need criteria. How about the gastrointestinal pathology textbooks? They have illustrations of LGD and HGD that are clearly entirely different. But discriminating between idealized LGD and HGD isn't the problem. No book gives us clear criteria for distinguishing the high-side of LGD from the low-side of HGD, which is what we really need to know for management purposes.
The literature is loaded with publications covering dysplasias in Barrett mucosa; presumably, the authors of these studies know how to tell LGD from HGD, and we, the readers, assume that written criteria exist. Here are a few examples of what the literature has to offer for diagnosing LGD and HGD. In a study of a large series of Barrett dysplasias that is frequently quoted in the gastroenterology literature,5 the authors discuss how they diagnosed the dysplasias. They state that "the interpretation of the microscopic findings was based on previously published definitions" and they refer to an article by Riddell et al6 that was published in 1983 and dealt not with Barrett mucosa, including dysplasias, but with colitis, mainly ulcerative colitis and colitic dysplasias. Barrett mucosa was nowhere mentioned in that article, yet it serves as a reference for a study of Barrett dysplasias. Furthermore, the cited article did not give clear diagnostic criteria for colitic dysplasias, only suggestions on how to make the diagnoses. In a study of LGD, Barrett biopsies were classified using previously published but modified criteria, with the same article by Riddell et al quoted as the source. The authors7 did not tell the reader what the modifications were or what characterized the highest end of LGD and separated it from HGD.
In a study by Montgomery and a group of pathologists specializing in gastrointestinal diseases,8 the pathologists exchanged slides of Barrett biopsies and classified the observed changes independently. The participants were said to be "familiar with the criteria proposed in 1988" for the epithelial changes in Barrett mucosa, "and they used these as a general basis for their interpretations." The article they cited was an earlier slide review study by Reid and a different group of 8 pathologists with gastrointestinal expertise (9) that looked at reproducibility of the diagnosis of dysplasia in Barrett mucosa. In the article, the features of both LGD and HGD were separated into architectural and nuclear or cytologic types, but these features were those of dysplasia in general, not specifically of a particular grade. Architectural changes included "budded, branched, crowded, or irregularly shaped glands, papillary extensions into gland lumina, and villiform configuration of the surface." Nuclear or cytologic changes included "marked variation in size and shape, nuclear and/or nucleolar enlargement, increased nuclear to cytoplasmic ratio, and hyperchromatism, as well as an increased numbers of abnormal mitoses." This last criterion is peculiar since it suggests that there are a normal number of abnormal mitoses. The article by Reid et al (9) also stated that "[t]he diagnosis of LGD or HGD is based on the severity of both architectural and cytologic criteria.... Although either architectural or cytologic abnormalities may predominate, HGD is diagnosed if either one is sufficiently prominent." This basically means that not all changes associated with one grade of dysplasia are present to the same extent. Thus the criteria in the article by Reid et al9 include changes that are sufficiently prominent. However it does not specify how prominent is sufficiently prominent. The answer is simple, it's more than insufficiently. So this paper, although it has been referenced many times, doesn't give us the information that we need.
Montgomery and her colleagues8 refined the criteria by Reid et al9 and stated that the diagnosis of HGD included these features: "... as in LGD, surface maturation is lacking. The architecture may show crowding of cytologically abnormal glands or be markedly distorted with glandular crowding and little lamina propria. If the cytologic features are sufficiently dysplastic, lesser architectural distortion is acceptable." The definition of HGD also mentions that "nuclei are hyperchromatic and nuclear membranes irregular. Cells may have either delicately clumped dark heterochromatin and inconspicuous nucleoli or prominent irregular nuclei with irregularly clumped chromatin and nucleoli or prominent irregular nuclei with irregular clumped chromatin and irregular nucleoli. Markedly enlarged hyperchromatic cells are a feature of HGD, and these may extend to the surface. Loss of nuclear polarity is seen. Mitoses are readily identifiable." This definition includes words such as "may," "sufficiently," "delicately," "inconspicuous," "prominent," "irregular," "irregularly," "abnormal," "markedly," and "readily identified." How do we tell marked from less than marked? Where does regular end and irregular begin? How abnormal is abnormal? How quickly is something readily identified? Besides, anything with the word "may" is clearly not a "hard" criterion. Other groups followed the approach of Montgomery and colleagues. As an example, in a HGD study from Buttar et al,10 "HGD was defined by criteria originally established by Reid and colleagues," but as mentioned above, the article by Reid et al9 did not establish criteria specifically for high-grade dysplasia.10
So how do we really tell the difference between HGD and LGD, since there are no specific rules, just suggestions? From a practical standpoint, we, pathologists, look at the epithelium, decide if it is dysplastic, and then try to determine if it looks "bad enough" to be HGD. If we have doubts, we show it to other pathologists and see if it looks bad enough to them. The minimal criteria exist in our eyeballs and in our brains. We all have different eyeballs, and we all have different brains, and that is why we have so much trouble telling one biopsy from another. Furthermore, as far as I know, there are no reproducibility studies on the diagnoses of the lowest end of HGD and the highest end of LGD. What we really need is a HGD stain or some molecular or genetic marker that has proven prognostic significance, independent of our eyeballs.
What changes in biopsy specimens lead to the diagnosis of invasive carcinoma? We have a better handle on this issue. Invasion into the lamina propria seems to be characterized by single neoplastic cells or small clusters of cells in the lamina propria, separated from the dysplastic tubules. Complex architecture, including solid areas of dysplastic cells or cribriform growth pattern and dysplastic cells incorporated within the squamous epithelium, are changes with a very high predictive value for invasive carcinoma in the lamina propria. (11) Desmoplasia surrounding dysplastic tubules is an excellent indication of invasive carcinoma, but desmoplasia occurs only when the cancer invades deeper than the lamina propria. It will not be found when the carcinomas are limited to the lamina propria.
Endoscopic mucosal resection specimens of Barrett mucosa have an anatomic peculiarity, that is, a double muscularis mucosae. This was first reported in 1981 by Kato et al, (12) from Japan, where the incidence of Barrett mucosa is small. Subsequently, in 1991, in a controlled study by Takubo et al, (13) also from Japan, the authors found that the double muscularis mucosae only occurred where there was Barrett mucosa; it occurred in 7 of the 8 cases of Barrett mucosa under study. They also found that the deeper of the 2 layers of muscularis mucosae was continuous with the normal esophageal muscularis mucosae. The inner layer was the duplication. The outer muscularis mucosae tends to be much thicker than the inner layer, and it can be confused with the muscularis propria on EMR specimens. Pathologists who deal with EMR specimens need to be aware of the duplicated muscularis mucosae for accurate staging, so as not to confuse this with the muscularis propria. (14)
The presence of a duplicated muscularis mucosae also means that there are 2 lamina proprias, 1 mixed with the epithelium toward the lumen and a second, deeper lamina propria between the 2 layers of muscularis mucosae. Furthermore, in many cases of Barrett mucosa, there is abundant smooth muscle and collagen in the superficial layer of lamina propria, probably part of the duplicated inner muscularis mucosae. This has been called the musculo-fibrous anomaly.15 These anatomic alterations suggest that Barrett mucosa is not just simply an epithelial aberration, but it is a complex abnormality involving both metaplastic epithelium and abnormal stroma. What is the prognosis when Barrett cancers invade into this deep lamina propria? Is it the same as invasion only into the upper lamina propria--in other words, the 2 layers of lamina propria are really the same--or is this deeper lamina propria the equivalent of submucosa, which makes a carcinoma within it much more likely to metastasize? This has been dealt with in a few recent publications. (16,17) Abraham et al (16) found that the risk of metastasis for invasion into any part of this complex, from the superficial lamina propria through the deep or original muscularis mucosae, is much the same, about 10%, much less than the risk of submucosal invasion, where metastases occur in about 30% of carcinomas.
Finally, in endoscopic mucosal resections for HGD and/or superficially invasive carcinomas that are near the gastroesophageal junction, what type of mucosa is actually the origin of these cancers? There are some recent data from Japan and Germany that cover a large number of EMR specimens with cancers, and the most common marginal mucosa was not Barrett, but was cardiac. (18) We have seen a number of EMRs in endoscopically suspected Barrett mucosa with no evidence of Barrett mucosa at all, just cardiac mucosa. It is possible that Barrett mucosa had been previously present but was replaced in toto by dysplastic and carcinomatous mucosa. Or perhaps Barrett mucosa was never present and the neoplastic epithelium developed within another type of epithelium. What does that mean in terms of surveillance for Barrett mucosa? Is it just Barrett mucosa that we need to be concerned about, or should we develop surveillance programs for other kinds of mucosae? Currently we don't know the answer.
[FIGURE 2 OMITTED]
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
[FIGURE 7 OMITTED]
CONTINUING WITH THE REPORT OF THE CASE
In addition to HGD, the EMR specimen had invasive adenocarcinoma, with small clusters of neoplastic cells in the lamina propria, complex architecture with interconnecting cords of cells and tubules, and incorporation of neoplastic cells into squamous mucosa (Figures 2-1). The carcinoma invaded no deeper than the muscularis mucosae (Figure 5). What about the margins of resection? The columnar margin was cauterized and HGD seemed to be present at this margin, but maybe there was invasive cancer as well at this margin that was obliterated by the cautery (Figure 6). There was no Barrett mucosa but there was a duplicated muscularis mucosae, strong evidence that Barrett mucosa had been present (Figure 7). This suggests that the dysplasia and carcinoma totally replaced the preexisting Barrett mucosa. The carcinoma's deepest level of invasion was into the superficial layer of muscularis mucosae, the new or duplicated layer, which carries a metastatic risk of perhaps 10%.
The final diagnosis is adenocarcinoma invading into the inner or duplicated muscularis mucosae, in the distal esophagus, probably arising in Barrett mucosa, on the basis of the double muscularis mucosae. The deep margin of resection is free, but dysplasia extends to a lateral margin. Cautery artifact at that margin damaged the tissues so severely that it was impossible to tell if, in addition to dysplasia, carcinoma also exists at that margin.
(1.) Wang KK, Sampliner RE. Updated guidelines 2008 for the diagnosis, surveillance and therapy of Barrett's esophagus. Am J Gastroenterol. 2008;103(3): 788-797.
(2.) Mino-Kenudson M, Brugge WR, Puricelli WP, et al. Management of superficial Barrett's epithelium-related neoplasms by endoscopic mucosal resection. Am J Surg Pathol. 2005;29(5):680-686
(3.) Prasad GA, Buttar NS, Wongkeesong LM, et al. Significance of neoplastic involvement of margins obtained by endoscopic mucosal resection in Barrett's esophagus. Am J Gastroenterol. 2007;102(11):2380-2386.
(4.) Larghi A, Lightdale CJ, Ross AS, et al. Long-term follow-up of complete Barrett's eradication endoscopic mucosal resection (CBE-EMR) for the treatment of high grade dysplasia and intramucosal carcinoma. Endoscopy. 2007;39(12): 1086-1091.
(5.) Snell TG, Sontag SJ, Chejfec G, et al. Long-term nonsurgical management of Barrett's esophagus with high-grade dysplasia. Gastroenterol. 2001;120(7):1607-1619.
(6.) Riddell RH, Goldman H, Ransohoff DF, et al. Dysplasia in inflammatory bowel disease: standardized classification with provisional clinical applications. Hum Pathol. 1983;14(11):931-968.
(7.) Skacel M, Petras RE, Gramlich TL, et al. The diagnosis of low-grade dysplasia in Barrett's esophagus and its implications for disease progression. Am J Gastroenterol. 2000;95(12):3383-3387.
(8.) Montgomery E, Bronner MP, Goldblum JR, et al. Reproducibility of the diagnosis of dysplasia in Barrett esophagus: a reaffirmation. Hum Pathol. 2001; 32(4):368-378.
(9.) Reid BJ, Haggitt RC, Rubin CE, et al. Observer variation in the diagnosis of dysplasia in Barrett's esophagus. Hum Pathol. 1988(2);19:166-178.
(10.) Buttar NS, Wang KK, Sebo TJ, et al. Extent of high-grade dysplasia in Barrett's esophagus correlates with risk of adenocarcinoma. Gastroenterol. 2001(7);120:1630-1639.
(11.) Zhu W, Appelman HD, Greenson, JK, et al. Barrett's/cardiac high grade dysplasia is not a strong marker for concurrent carcinoma, unless architectural changes suspicious for adenocarcinoma are also present. Mod Pathol. 2006; 19(suppl 1):126A.
(12.) Kato H, Iizuka T, Watanabe H, et al. Double adenocarcinoma in Barrett's esophagus. Jpn J Clin Oncol. 1981;11(4):523-530.
(13.) Takubo K, Sasajima K, Yamashita K, et al. Double muscularis mucosae in Barrett's esophagus. Hum Pathol. 1991;22(11):1158-1161.
(14.) Lewis JT, Wang KK, Abraham SC. Muscularis mucosae duplication and the musculo-fibrous anomaly in endoscopic mucosal resections for Barrett esophagus: implications for staging of adenocarcinoma. Am J Surg Pathol. 2008;32(4): 566-571.
(15.) Rubio CA, Riddell R. Musculo-fibrous anomaly in Barrett's mucosa with dysplasia. Am J Surg Pathol. 1988;12(11):885-886.
(16.) Abraham SC, Krasinskas AM, Correa AM, 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(11):1719-1725.
(17.) van Sandick JW, van Lanschot JJB, ten Kate FJW, et al. Pathology of early invasive adenocarcinoma of the esophagus or esophagogastric junction: implications for therapeutic decision making. Cancer. 2000;88(11):2429-2437.
(18.) Takubo K, Aida J, Naomoto Y, et al. Cardiac rather than intestinal-type background in endoscopic resection specimens of minute Barrett adenocarcinoma. Hum Pathol. 2009;40(1):65-74.
Henry D. Appelman, MD
Accepted for publication June 22, 2009.
From the Department of Pathology, University of Michigan, Ann Arbor.
The author has no relevant financial interest in the products or companies described in this article.
Presented at New Frontiers in Pathology: An Update for Practicing Pathologists, University of Michigan, Ann Arbor, September 19, 2008.
Reprints: Henry D. Appelman, MD, Department of Pathology, University of Michigan, 1500 E Medical Center Dr, Room 2G332, Ann Arbor, MI 48109 (e-mail: email@example.com).
|Printer friendly Cite/link Email Feedback|
|Author:||Appelman, Henry D.|
|Publication:||Archives of Pathology & Laboratory Medicine|
|Article Type:||Clinical report|
|Date:||Nov 1, 2009|
|Previous Article:||Common diagnostic challenges in the pathology of nonneoplastic lung diseases: a case-based review.|
|Next Article:||Association of cystic neck metastases and human papillomavirus-positive oropharyngeal squamous cell carcinoma.|