Adenocarcinoma of the distal esophagus is a recognized

ORIGINAL ARTICLE Adenocarcinomas of the Distal Esophagus and Gastric Cardia Are Predominantly Esophageal Carcinomas Parakrama Chandrasoma, MD, Kumari Wickramasinghe, MD, PhD, Yanling Ma, MD, and Tom DeMeester, MD Background: Adenocarcinoma of the distal esophagus and gastric cardia are defined by the relationship of its epicenter to the gastro-esophageal junction, which is presently defined as the end of the tubular esophagus. We have recently suggested that the true gastro-esophageal junction is best defined by the proximal limit of gastric oxyntic mucosa. Aim: To reclassify adenocarcinomas of this region by the relationship of the tumor to the proximal limit of gastric oxyntic mucosa. Methods: Seventy-four patients who had esophago-gastrectomy for adenocarcinomas in this region were classified as adenocarcinoma of distal esophagus (38 patients) and gastric cardia (36 patients) by present criteria. The epithelial type at the epicenter and distal edge of these tumors was assessed. Results: The epicenter of the tumor in 64 patients with noncircumferential tumors had squamous (5 cases), cardiac (21 cases), oxynto-cardiac (4 cases), and intestinal (Barrett-type) (34 cases) epithelia. None had gastric oxyntic mucosa. Of the 10 patients with circumferential tumors, 7 had cardiac or oxyntocardiac epithelium at the distal tumor edge. Conclusions: If the gastro-esophageal junction is defined histologically as the proximal limit of oxyntic mucosa, 71/74 patients would be classified as adenocarcinoma of the distal esophagus. The other 3 patients were questionable as to gastric or esophageal origin. We suggest that this reclassification based on the proposed new definition of the gastro-esophageal junction provides an explanation for the epidemiologic relationship that exists between adenocarcinoma of the gastric cardia and gastro-esophageal reflux disease. Key Words: gastro-esophageal junction, esophagus, adenocarcinoma, gastric cardia, Barrett esophagus (Am J Surg Pathol 2007;31:569 575) From the Departments of Surgical Pathology and Foregut Surgery, Keck School of Medicine and University of Southern California, Los Angeles, CA. Supported by any external funding source. Reprints: Parakrama Chandrasoma, MD, Department of Surgical Pathology, LAC+USC Medical Center, GH 16-905, 1200 N. State Street, Los Angeles, CA 90033 (e-mail: ptchandr@usc.edu). Copyright r 2007 by Lippincott Williams & Wilkins Adenocarcinoma of the distal esophagus is a recognized complication of gastro-esophageal reflux. 15 Its incidence has increased 6-fold in the past 3 decades in the United States and Western Europe. 4,5,13,18,19,20 A similar increase has been shown in the incidence of adenocarcinoma of the gastric cardia 4,5,13,20 with a recent trend suggesting a flattening of the incidence curve for this location that has not been seen for esophageal adenocarcinoma. 19 Somewhat surprisingly, epidemiologic studies have shown that there is a correlation between adenocarcinoma of the gastric cardia and symptomatic gastro-esophageal reflux, although this correlation is weaker than that for adenocarcinoma of the distal esophagus. 15 Classifying tumors of this region into distal esophageal and gastric cardiac is based on endoscopic and radiologic criteria 15,24,25 or pathologic examination of resected specimens. 10 Endoscopically, the gastroesophageal junction is defined as the proximal limit of rugal folds. 16,22 In 2000, the Association of Directors of Anatomic and Surgical Pathology (ADASP) made recommendations for standardizing the classification of these tumors in resected specimens. 2 They advocated the use of the relationship of the epicenter of the tumor to the grossly defined gastro-esophageal junction. This group of experts defined the gastro-esophageal junction as the horizontal line drawn across the end of the tubular esophagus. When classified by the present definitions of the gastro-esophageal junction, there is evidence that gastric cardiac adenocarcinomas commonly arise in intestinal metaplasia of the gastric cardia, 21 and that they bear similarities to distal esophageal adenocarcinoma. 6 We have suggested that the gastro-esophageal junction is most accurately defined by histologic determination of the proximal limit of gastric oxyntic mucosa. 7 10 The area proximal to this, which is esophagus, is lined by squamous epithelium in normal people and a variable extent of metaplastic esophageal columnar epithelium (oxynto-cardiac, cardiac, and intestinal) in patients with reflux damage. In our recent study of 10 esophagectomy specimens, 10 we showed that there was a region between the end of the tubular esophagus (the present definition of the gastro-esophageal junction) and the proximal limit of gastric oxyntic mucosa (the true gastro-esophageal junction) that was identified as esophagus rather than stomach by the presence of submucosal esophageal glands (Fig. 1). We designated this area as the dilated Am J Surg Pathol Volume 31, Number 4, April 2007 569

Chandrasoma et al Am J Surg Pathol Volume 31, Number 4, April 2007 FIGURE 1. Definition of the dilated end-stage esophagus, which is the area distal to the end of the tubular esophagus (the present definition of the gastro-esophageal junction) and the proximal limit of gastric oxyntic mucosa (the true gastroesophageal junction) (gray = squamous epithelium; black = cardiac mucosa with and without intestinal metaplasia; white = gastric oxyntic mucosa. The submucosal glands are shown as yellow circles). end-stage esophagus. The use of the end of the tubular esophagus resulted in a discrepancy of 0.31 to 2.05 cm between what is presently called the gastro-esophageal junction and what we have shown is the true gastroesophageal junction. 10 Sarbia et al, 23 in a similar study of esophagectomy specimens, showed that cardiac and oxynto-cardiac mucosa can extend to 2.8 cm beyond the end of the tubular esophagus. This error would have the potential of distal esophageal adenocarcinomas being incorrectly designated as gastric cardiac adenocarcinomas. We undertook this study to compare how adenocarcinomas of this region would be classified on the basis of the presently accepted gross definition of the gastroesophageal junction for use in resected specimens and the new histologic definition of the gastro-esophageal junction that we have proposed. We were unable to test these tumors against the presently accepted endoscopic definition of the gastro-esophageal junction; in many of these cases, the rugal folds are greatly distorted in this region by the presence of the tumor. FIGURE 2. Adenocarcinoma of the distal esophagus by present definition. The tumor extends focally across the line of the end of the tubular esophagus but has its epicenter slightly above the line. to the tumor, and at the lateral edge of the epicenter of the tumor. The last determination was not possible in 10 cases where the tumor involved the full circumference of the esophagus. MATERIALS AND METHODS Seventy-four patients who underwent esophagogastrectomy for adenocarcinoma of the distal esophagus and gastric cardia (proximal stomach) during the years of 1997 to 2000 were selected for study. The relationship of the epicenter of the tumor to the grossly defined gastroesophageal junction was used to classify these tumors as distal esophageal (Fig. 2) and gastric cardiac (Fig. 3). The gastro-esophageal junction was defined as a horizontal line drawn between the end of the tubular esophagus and the saccular stomach as recommended by the ADASP. 2 The specimen was sectioned extensively in a manner that permitted evaluation of the mucosal types present immediately proximal to the tumor, immediately distal FIGURE 3. Adenocarcinoma of the gastric cardia, by present definition. The tumor presents as an ulcerated mass whose epicenter is clearly below the end of the tubular esophagus. 570 r 2007 Lippincott Williams & Wilkins

Am J Surg Pathol Volume 31, Number 4, April 2007 Origin of Adenocarcinomas of the GEJ Region FIGURE 4. Cardiac mucosa, consisting entirely of mucous cells in surface, foveolar region and glands. Parietal and goblet cells are absent. The epithelial types were classified by previous reported criteria 7,9 into squamous, cardiac (epithelium composed only of mucous cells without any parietal cells, Fig. 4), oxynto-cardiac (containing glands composed of a mixture of parietal and mucous cells, Fig. 5), intestinal (defined by the presence of goblet cells in cardiac mucosa, Fig. 6), and oxyntic (containing glands composed of parietal and chief cells without mucous cells, Fig. 7). The type of epithelium at the proximal, distal, and central lateral edge of each tumor was recorded. Where more than one epithelial type was present, priority was given as follows: intestinal, cardiac, oxynto-cardiac, squamous, and oxyntic, on the basis of our belief that the first 3 represents the gradation of these epithelial types in this region from most to least abnormal of reflux-induced FIGURE 6. Cardiac mucosa with intestinal metaplasia, characterized by the presence of well-defined goblet cells on hematoxylin and eosin-stained section. No parietal cells are seen. columnar epithelia of the esophagus, and the last 2 are normal epithelia of esophagus and stomach, respectively. 8 RESULTS There were 58 men and 16 women in this series for a M:F ratio of 3.6:1. By the present criterion for the gastroesophageal junction (a line drawn across the end of the tubular esophagus), 38 (51%) of these tumors were classified as distal esophageal (Fig. 2) and 36 (49%) as gastric cardiac (Fig. 3). Of the 38 distal esophageal cancers, 30 were men and 8 (21%) were women. Eight FIGURE 5. Oxynto-cardiac mucosa, showing a mixture of parietal cells and mucous cells in glands under the foveolar region. Goblet cells are absent. FIGURE 7. Gastric oxyntic mucosa with only parietal and chief cells without mucous cells in the glands below the foveolar region. r 2007 Lippincott Williams & Wilkins 571

Chandrasoma et al Am J Surg Pathol Volume 31, Number 4, April 2007 (25%) of the 36 patients with gastric cardiac tumors were women. The mean age was 64.1 years (median: 66 y; range: 31 to 86 y). The tumor size ranged from grossly invisible carcinomas to large tumors exceeding 15 cm in greatest dimension. Histologically, they were all pure adenocarcinomas and ranged from well differentiated to poorly differentiated and showed many different histologic subtypes including tubular, mucinous, papillary, signet ring cell, solid, microcystic, and clear cell types of adenocarcinoma. The depth of invasion of the tumor was intramucosal in 14 (18.9%), submucosal in 8 (10.8%), intramural in 4 (5.4%), and transmural in 48 cases (64.9%). Helicobacter pylori gastritis was present in ten of 74 (14%) patients which is similar to the reported prevalence of H. pylori infection in our previous study of carditis in our study population. 12 These 10 patients included 5 of 38 with distal esophageal tumors and 5 of 25 with gastric cardiac tumors. Only one of these, a patient with a distal esophageal tumor, had multifocal atrophic gastritis with intestinal metaplasia in the gastric part of their resection specimen. This patient had cardiac mucosa without intestinal metaplasia at the distal edge of the tumor. Lymph node involvement correlated with depth of tumor invasion. There was 1 patient among the 14 intramucosal tumors that was node positive; this patient had a poorly differentiated intramucosal carcinoma that measured 3 cm with 1 positive node out of 45 nodes. Two (25%) of 8 patients with submucosal tumors, 3 (75%) of 4 patients with intramural tumors, and 43 (90%) of 48 patients with transmural tumors had positive lymph nodes. The epithelial type at the proximal and distal edges of the tumor was determined by examination of the section immediately above the upper edge and below the distal edge of the tumor, respectively (Table 1). Of interest is the fact that 25 patients had nondysplastic intestinal metaplasia proximal to the tumor, suggesting that the carcinomas arose in the more distal part of the segment of Barrett esophagus. No tumors had oxynto-cardiac or oxyntic mucosa proximal to the tumor. The epithelial type at the lateral edge of the epicenter of the tumor was determined by examination of the section across the center of the tumor in the 64 cases where taking such a section was feasible (Fig. 8). The epithelial type at the center of the tumor was squamous in 5 patients, intestinal in 34 (53%) patients, cardiac in 21 (33%) patients, and oxynto-cardiac in 4 patients. None of the 64 noncircumferential tumors had oxyntic mucosa at the epicenter (Table 1). In 10 patients, circumferential involvement of the esophagus precluded the evaluation of the epithelial type at the lateral edge of the epicenter of the tumor. These included 4 distal esophageal tumors and 6 gastric cardiac tumors. Examination of the epithelium distal to the distal limit of these tumors showed cardiac mucosa in 3 cases and oxynto-cardiac mucosa in 4 cases; the other 3 patients had oxyntic mucosa at the distal tumor edge. These 3 tumors were all large (4, 6.5, and 11 cm), with transmural TABLE 1. Epithelial Types Found at Proximal Edge, Distal Edge, and Lateral Edge at the Epicenter of Distal Esophageal and Gastric Cardiac Adenocarcinomas Classified by its Relationship to the end of the Tubular Esophagus Epithelial Type Distal Esophageal Tumors (n = 38) Gastric Cardiac Tumors (n = 36) At proximal edge of tumor Squamous 16 (42%) 27 (75%) Intestinal 20 (53%) 5 (14%) Cardiac 2 (5%) 4 (11%) Oxynto-cardiac 0 0 Oxyntic 0 0 At the lateral edge of epicenter of tumor Squamous 2 (5%) 3 (8%) Intestinal 25 (66%) 9 (25%) Cardiac 5 (13%) 16 (44%) Oxynto-cardiac 2 (5%) 2 (6%) Oxyntic 0 0 Circumferential 4 (11%) 6 (17%) tumor At the distal edge of tumor Squamous 0 0 Intestinal 12 (32%) 4 (11%) Cardiac 13 (34%) 14 (39%) Oxynto-cardiac 7 (18%) 8 (22%) Oxyntic 6 (16%) 10 (28%) invasion and positive lymph nodes (17 of 62, 8 of 26, and 9 of 36 lymph nodes positive), lymphovascular invasion, undermining of proximal squamous epithelium and distal oxyntic mucosa and with satellite nodules in the mucosa away from the main tumor. The epithelial type at the distal edge of the tumor was determined by examination of the vertical section taken across the distal edge of the tumor and are shown in Table 1. Sixteen patients, which included 6 distal esophageal tumors and 10 gastric cardiac tumors, had gastric oxyntic mucosa at the distal tumor edge. These were characterized by being large tumors (size range from 2 to 11 cm; mean size 6 cm; median size 5.5 cm), deeply FIGURE 8. Lateral edge at the epicenter of a moderately differentiated adenocarcinoma of the gastric cardia showing cardiac mucosa at the edge of the tumor. 572 r 2007 Lippincott Williams & Wilkins

Am J Surg Pathol Volume 31, Number 4, April 2007 Origin of Adenocarcinomas of the GEJ Region invasive (all 16 tumors were transmural) with evidence of lymph node involvement (all 16 tumors showed lymph node involvement). These tumors also showed frequent undermining of the proximal squamous epithelium (15 cases), undermining of distal oxyntic mucosa (10 cases), lymphovascular involvement (14 cases), and the presence of satellite tumor nodules in squamous or oxyntic mucosa (12 cases). DISCUSSION In 1953, Allison and Johnstone 1 described the columnar lined esophagus interposed between the squamocolumnar junction and the gastro-esophageal junction. Before this, columnar lined esophagus was called a tubular intrathoracic stomach by virtue of the fact that the esophagus was believed to end at the squamocolumnar junction. 3 In a masterful dissection of one of their resection specimen, the pathologist, Dr D. H. Collins, proved that what was being called tubular stomach was in fact esophagus. 1 He reports: The esophagus was separated with a knife from the stomach along the line of peritoneal reflection. A vertical slice was then made through the centre of the reconstituted specimenyy, and three vertically contiguous blocks were preparedy The histologic examination is reported: The stomach below the anatomical junction with the esophagus is lined by gastric mucosa of fundal typey.. Cardiac glands and cardiac gastric mucosa do not appear until 0.6 cm up the anatomic esophagus (Fig. 9). Allison and Johnstone 1 identify the features that prove that the columnar lined structure is esophagus rather than stomach: it has no peritoneal covering, that the musculature is that of the normal esophagus, that there may be islands of squamous epithelium within it, that there are no oxyntic cells in the mucosa, and that in addition to gastric glands there are present typical esophageal mucous glands (Fig. 9). When the validated definition of the true gastroesophageal junction (the peritoneal reflection) is used, there is no confusion. The epithelium at and distal to the gastro-esophageal junction is gastric fundal ( = oxyntic) mucosa. The columnar lined esophagus is defined by an epithelium that is largely cardiac. This is true whenever a validated definition of the true gastro-esophageal junction is used. De Hertogh et al, 11 in a study of fetal specimens showed that the epithelium at the gastro-esophageal junction (the angle of His) contained parietal cells. Paull et al, 17 in a study of columnar lined esophagus showed that the epithelium in and above the upper end of the lower esophageal sphincter contained parietal cells. We have shown that the use of the end of the tubular esophagus or proximal limit of the rugal folds to define the gastro-esophageal junction place the gastroesophageal junction at a point that can be over 2 cm proximal to the true gastro-esophageal junction. 10 The effect of this error is to mis-label the most distal part of the columnar lined esophagus as the proximal stomach. This area that is presently mis-labeled proximal stomach FIGURE 9. Diagrammatic representation of the dissection in Allison and Johnstone of their case 1 by the pathologist, Dr Collins. The epithelium at the peritoneal reflection (the true gastro-esophageal junction) was gastric oxyntic mucosa. Submucosal glands were present in the nonoxyntic mucosa above this line that defined the columnar lined esophagus for the first time (gray = squamous epithelium; black = cardiac mucosa with and without intestinal metaplasia; white = gastric oxyntic mucosa. The submucosal glands are shown as yellow circles). can be recognized as esophagus by the fact that it is not covered by a peritoneal lining, it sometimes contains squamous islands within it, 14 it is proximal to gastric oxyntic mucosa and lined by cardiac mucosa (with and without intestinal metaplasia) and oxynto-cardiac mucosa identical to that seen in the columnar lined esophagus, and it contains submucous esophageal glands. 10 These are the same criteria that Allison and Johnstone 1 used to convince Norman Barrett that what he was calling the tubular stomach was actually esophagus. We should recognize that this is true today as well: these same criteria should convince the reader that what we are calling proximal stomach today is actually esophagus when it displays these criteria. If this is true, pathologic entities that are presently classified as lesions of the gastric cardia, potentially demand reclassification as esophageal lesions. These include adenocarcinoma of the gastric cardia and intestinal metaplasia of the gastric cardia. The 2 entities are troublesome entities at present because, although they are classified as gastric diseases, they have strong evidence to indicate that they are associated with gastro-esophageal reflux disease. It is difficult to understand how gastro-esophageal reflux disease can result in gastric pathology. Also, adenocarcinoma of the gastric cardia has epidemiologic similarities to esophageal rather than gastric adenocarcinoma. Our division of these tumors in the present study into adenocarcinoma of the esophagus and gastric cardia conforms to the International Classification of Diseases. In this classification, tumors of the gastro-esophageal junction are classified within the definition of gastric r 2007 Lippincott Williams & Wilkins 573

Chandrasoma et al Am J Surg Pathol Volume 31, Number 4, April 2007 cardiac tumors. When so classified, 38 patients had distal esophageal tumors and 36 had gastric cardiac tumors. We applied the same recommendation presently recommended by the ADASP to classify the tumor by the relationship of its epicenter to the newly proposed gastroesophageal junction (the histologically defined proximal limit of gastric oxyntic mucosa). The epithelium at the epicenter of the tumor was defined by a lateral section across the center of the tumor. If the epithelium at the epicenter of the tumor was gastric oxyntic mucosa, the tumor was classified as gastric; if it was squamous or cardiac mucosa (with and without intestinal metaplasia) and oxynto-cardiac mucosa, the tumor was classified as esophageal. In 10 patients, this was not possible because the tumor was circumferential. When so reclassified, all 64 tumors (34 distal esophageal and 30 proximal gastric) that were noncircumferential had cardiac mucosa (with and without intestinal metaplasia) and oxynto-cardiac mucosa at their epicenter and were classified as esophageal. This meant that 30 of 36 tumors previously called gastric cardiac in location were now reclassified as esophageal tumors. The use of the relationship of the epicenter of a tumor to the gastro-esophageal line to classify tumors of this region, as suggested by most experts, is reasonable. It, however, assumes equivalent growth of the tumor proximally and distally. For example, if a tumor arises 1 mm above the gastro-esophageal junction and infiltrates distally into the stomach in preference to proximally into the esophagus, it is liable to be mis-classified as a gastric cancer when it becomes large. A more specific criterion for classification would be the relationship of the distal edge of the tumor to the gastro-esophageal line. If the distal edge of the tumor was proximal to the gastroesophageal line, there can be no argument about the esophageal derivation of the tumor. In the present study, the distal edge of the tumor showed oxyntic mucosa only in 16 patients; in the other 58 patients, the mucosa distal to the tumor was metaplastic esophageal epithelia (oxynto-cardiac in 15 patients, cardiac in 27 patients and intestinal in 16 patients). By our new proposed definition of the proximal limit of oxyntic mucosa as the true gastro-esophageal junction, the entire tumor in these 58 patients was esophageal. These 58 patients unquestionably have esophageal cancers because no part of their tumor is in the stomach. Ten patients had circumferential tumors, precluding assessment of the epithelium at the epicenter of the tumor. Examination of the epithelium at the distal edge of these 10 circumferential tumors showed metaplastic glandular epithelium in 7 patients (cardiac mucosa in 3 cases and oxynto-cardiac mucosa in 4 cases). In these 7 cases, therefore, the entire tumor was proximal to the proximal limit of oxyntic mucosa, and therefore esophageal. Three patients had oxyntic mucosa at the distal tumor edge; these tumors could be either true gastric adenocarcinomas or esophageal adenocarcinomas extending into the stomach distally. The 16 patients who had oxyntic mucosa at the distal edge of the tumor may have an adenocarcinoma of the esophagus infiltrating across the gastro-esophageal junction to involve gastric oxyntic mucosa or they may have gastric adenocarcinoma infiltrating the esophagus. The epicenter in 13 of these tumors that were noncircumferential showed squamous (3 cases), cardiac (6 cases), oxynto-cardiac (1 case), or intestinal (3 cases) epithelia. This indicates that in these 13 cases more than half the tumor was in the esophagus but it does not prove that the origin was esophageal because of the possibility of nonuniform growth. By the present ADASP criterion of classifying these tumors to the relationship of the epicenter, 2 these tumors would also be esophageal. These tumors were all large (mean size 6 cm) and extensively infiltrating tumors. They all showed transmural invasion with evidence of lymph node involvement (all cases were node positive) and showed undermining of the proximal squamous epithelium (15 cases), undermining of distal oxyntic mucosa (13 cases), lymphovascular involvement (14 cases), and the presence of satellite tumor nodules (12 cases) in squamous lined and oxyntic mucosa. The 3 circumferential tumors with oxyntic mucosa distally also had all of these features. Using our new proposed criterion of the proximal limit of oxyntic mucosa as the true gastro-esophageal junction had the effect of shifting this line distally to the extent that 58 of these cancers were entirely in the esophagus. Thirteen of the other 16 tumors that had oxyntic mucosa at its distal edge would have been classified as esophageal if one applied the presently accepted standard of taking the epicenter of the tumor and its relationship to the gastro-esophageal junction to indicate origin. It is only in the 3 circumferential cases with oxyntic mucosa at the distal end that there is any reasonable doubt about origin from esophagus or stomach. The reclassification of these tumors is based on a newly proposed definition of the gastro-esophageal junction: the proximal limit of gastric oxyntic mucosa. We believe that this is more accurate than the presently used definitions because it has been validated by the criterion of submucosal esophageal glands. We suggest it is worthy of examination and attempted duplication by other workers because it provides an explanation for the anomaly that adenocarcinomas presently classified as gastric cardiac have an epidemiologic association with symptomatic gastro-esophageal reflux disease. This correction would result in the probability that the finding of intestinal metaplasia in cardiac mucosa distal to the presently accepted gastro-esophageal junction is equivalent to Barrett esophagus and therefore the precursor lesion for most tumors presently believed to be gastric cardiac. 11 Whether this means that surveillance is indicated when intestinal metaplasia is found in this region will depend on the magnitude of cancer risk associated with this finding. This is presently unknown. 574 r 2007 Lippincott Williams & Wilkins

Am J Surg Pathol Volume 31, Number 4, April 2007 Origin of Adenocarcinomas of the GEJ Region REFERENCES 1. Allison PR, Johnstone AS. The oesophagus lined with gastric mucous membrane. Thorax. 1953;8:87 101. 2. Association of Directors of Anatomic and Surgical Pathology. Recommendations for reporting of resected esophageal adenocarcinomas. Am J Surg Pathology. 2000;31:1188 1190. 3. Barrett NR. Chronic peptic ulcer of the oesophagus and oesophagitis. Br J Surg. 1950;38:175 182. 4. Blot WJ, Devesa SS, Kneller RW, et al. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA. 1991; 265:1287 1289. 5. Blot WJ, Devesa SS, Fraumeni JF. Continued climb in rates of esophageal adenocarcinoma: an update. JAMA. 1993;270:1320. 6. Cameron AJ. Epidemiology of columnar lined esophagus and adenocarcinoma. Gastroenterol Clin N Am. 1997;26:487 494. 7. Chandrasoma P. Controversies of the cardiac mucosa and Barrett s esophagus. Histopathol. 2005;46:361 373. 8. Chandrasoma PT, Der R, Ma Y, et al. Histology of the gastroesophageal junction: an autopsy study. Am J Surg Pathol. 2000;24:402 409. 9. Chandrasoma PT, Lokuhetty DM, DeMeester TR, et al. Definition of histopathologic changes in gastroesophageal reflux disease. Am J Surg Pathol. 2000;24:344 351. 10. Chandrasoma P, Makarewicz K, Wickramasinghe K, et al. A proposal for a new validated histologic definition of the gastroesophageal junction. Hum Pathol. 2006;37:40 47. 11. De Hertogh G, Van Eyken P, Ectors N, et al. On the existence and location of cardiac mucosa; an autopsy study in embryos, fetuses, and infants. Gut. 2003;52:791 796. 12. Der R, Tsao-Wei DD, DeMeester T, et al. Carditis. A manifestation of gastroesophageal reflux disease. Am J Surg Pathol. 2001;25: 245 252. 13. Devesa SS, Blot WJ, Fraumeni JF. Changing patterns in the incidence of esophageal and gastric carcinoma in the United States. Cancer. 1998;83:2049 2053. 14. Fass R, Sampliner RE. Extension of squamous epithelium into the proximal stomach: a newly recognized mucosal abnormality. Endoscopy. 2000;32:27 32. 15. Lagergren J, Bergstrom R, Lindgren A, et al. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med. 1999;340:825 831. 16. McClave SA, Boyce HW Jr, Gottfried MR. Early diagnosis of columnar lined esophagus: a new endoscopic diagnostic criterion. Gastrointest Endosc. 1987;33:413 416. 17. Paull A, Trier JS, Dalton MD, et al. The histologic spectrum of Barrett s esophagus. N Eng J Med. 1976;295:476 480. 18. Pera M, Cameron AJ, Trastek VF, et al. Increasing incidence of adenocarcinoma of the esophagus and esophagogastric junction. Gastroenterology. 1993;104:510 513. 19. Pohl H, Welch HG. The role of overdiagnosis and reclassification in the marked increase of esophageal adenocarcinoma incidence. J Natl Cancer Inst. 2005;97:142 146. 20. Powell J, McConkey CC. The rising trend in esophageal adenocarcinoma and gastric cardia. Eur J Cancer Prev. 1992;1: 265 269. 21. Ruol A, Parent A, Zaninotto G, et al. Intestinal metaplasia is the probable common precursor of adenocarcinoma in Barrett esophagus and adenocarcinoma of the gastric cardia. Cancer. 2000;88:2520 2528. 22. Sampliner RE. Practice guidelines on the diagnosis, surveillance, and therapy of Barrett s esophagus. Am J Gastroenterol. 1998;93: 1028 1031. 23. Sarbia M, Donner A, Gabbert HE. Histopathology of the gastroesophageal junction. A study on 36 operation specimens. Am J Surg Pathol. 2002;26:1207 1212. 24. Siewert JR, Stein HJ. Adenocarcinoma of the gastroesophageal junction. Classification, pathology and extent of resection. Dis Esoph. 1996;9:173 182. 25. Siewert JR, Stein HJ. Classification of adenocarcinoma of the oesophagogastric junction. Br J Surg. 1998;85:1457 1459. r 2007 Lippincott Williams & Wilkins 575