- Email: [email protected]
Specialized columnar epithelium of the esophagogastric junction: prevalence and associations
THE AMERICAN JOURNAL OF GASTROENTEROLOGY © 1999 by Am. Coll. of Gastroenterology Published by Elsevier Science Inc.
Vol. 94, No. 4, 1999 ISSN 0002-9270/99/$20.00 PII S0002-9270(99)00041-6
Specialized Columnar Epithelium of the Esophagogastric Junction: Prevalence and Associations Markku Voutilainen, M.D., Martti Fa¨rkkila¨, M.D., Matti Juhola, M.D., Kyo¨sti Nuorva, M.D., Kari Mauranen, M.Sc., Timo Ma¨ntynen, M.D., Ilkka Kunnamo, M.D., Jukka-Pekka Mecklin, M.D., Pentti Sipponen, M.D., and The Central Finland Endoscopy Study Group* Departments of Internal Medicine, Pathology, and Surgery, Jyva¨skyla¨ Central Hospital, Jyva¨skyla¨; Division of Gastroenterology, Department of Internal Medicine, Helsinki University Hospital, Helsinki; Department of Health Policy and Management, Kuopio University, Kuopio; Karstula Health Centre, Karstula; Department of Pathology, Jorvi Hospital, Espoo, Finland
OBJECTIVES: In Barrett’s esophagus (BE) normal squamous esophageal epithelium is replaced by specialized columnar epithelium (SCE). BE is related to gastroesophageal reflux disease (GERD) and is a risk factor for esophageal adenocarcinoma. SCE is detected also at normal-appearing esophagogastric junction without BE (junctional SCE). The relationships between junctional SCE, GERD, and cardia adenocarcinoma are obscure and controversial. The aims of the present study were to investigate the prevalence and demographics of junctional SCE and to compare these figures with those reported for BE, and esophageal and cardia adenocarcinoma. A further aim was to examine the association between junctional SCE and GERD, Helicobacter pylori infection, and gastritis. METHODS: One thousand one hundred-nineteen consecutive dyspeptic patients underwent gastroscopy and were enrolled into the study. RESULTS: Junctional SCE was detected in 110 patients (10%). The age-specific prevalence of junctional SCE increased with age. The male:female ratio was 1:1.1. In multivariate analysis, junctional SCE was independently and positively related to endoscopic erosive esophagitis (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.1–3.1), cardia inflammation (carditis) (OR, 3.1; 95% CI, 1.4 – 6.8), and age (OR, 1.4 per decade; 95% CI, 1.2–1.6), but not to corpus H. pylori infection (OR, 1.4; 95% CI, 0.7–2.8), antral (OR, 1.0; 95% CI, 0.5–2.1) or corpus (OR, 0.8; 95% CI, 0.4 –1.8) gastritis, or intestinal metaplasia of the antral mucosa in stomach (OR, 1.2; 95% CI, 0.7–2.1). In univariate analysis, junctional SCE was, however, significantly more common in patients with antral-predominant atrophic gas-
* The following investigators are members of The Central Finland Endoscopy Study Group: Liisa Ahlskog-Muraja, Teuvo Antikainen, Sirpa Antila, Jorma Anttinen, Matti Hallikas, Pekka Hannonen, Kari-Pekka Ha¨ma¨la¨inen, Heikki Janhonen, Matti Kairaluoma, Kerkko Karjalainen, Pekka Kauranen, Heikki Korhonen, Jari Korhonen, Ritva Koskela, Raimo Krees, Vesa Ka¨rja¨, Pa¨ivi Laaksonen, Matti Laukkanen, Reino Liisanantti, Marja Lohman, Antero Palmu, Ulla Palmu, Matti Pellinen, Pertti Sa¨rkka¨, Harri Sela¨nne, Tuula Tervo, Marianne Udd, Jukka Viinikka.
tritis (20%), compared with those with normal gastric histology (8%, p , 0.001). CONCLUSIONS: Junctional SCE is age related and may therefore be an acquired lesion. It is associated with cardia inflammation and endoscopic erosive esophagitis, but not with H. pylori infection or gastric intestinal metaplasia. Unlike BE and cardia cancer, junctional SCE occurs with similar frequency in men and women. (Am J Gastroenterol 1999;94:913–918. © 1999 by Am. Coll. of Gastroenterology)
INTRODUCTION The diagnosis of Barrett’s esophagus (BE) requires the detection of the replacement of normal esophageal stratified squamous epithelium by specialized columnar epithelium (SCE) without any specification of the length of the metaplastic segment (1, 2). BE is an acquired condition related to gastroesophageal reflux disease (GERD) and harbors an increased risk of esophageal adenocarcinoma (1, 3). In some patients SCE is detected at the normal-appearing gastroesophageal (GE) junction, without BE (3, 4). The biological significance of the latter SCE type is unknown, nor is it known whether SCE at the normal-appearing GE junction is the precursor lesion for BE and cardia adenocarcinoma. Specialized columnar epithelium is an incomplete form of intestinal metaplasia characterized by goblet cell metaplasia and histochemical evidence of acidic sialomucins and sulphomucins in goblet cells and adjacent gastric-appearing columnar cells (5). It is easily and unequivocally detected in gastric biopsies with such simple histochemical stains as alcian-blue pH 2.5/periodic acid-Schiff or high-iron diamine (5, 6). The first aim of this study was to investigate the prevalence and demography of the SCE lesion of the esophagogastric junction in consecutive outpatients referred to endoscopy for dyspeptic symptoms and to compare the demographics of junctional SCE with those reported previ-
914
Voutilainen et al.
AJG – Vol. 94, No. 4, 1999
ously for BE and cardia cancer. The second aim was to investigate factors, such as gastritis and gastroesophageal reflux disease, that may be associated with junctional SCE.
MATERIALS AND METHODS Patients The study population consisted of the 1698 dyspeptic patients sent for gastroscopy between September 1 and December 31, 1996 in the Jyva¨skyla¨ Central Hospital referral area, the population of which is approximately 260,000 inhabitants. Adequate esophagogastric junctional biopsy specimens were obtained from 1184 (69%) patients. The patients with long- or short-segment Barrett’s esophagus or SCE detected in the esophagus without any specification of the length of the metaplastic segment, previous H. pylori eradication, and gastric surgery were excluded. Thus the final study population consisted of 1119 patients (516 men and 603 women). Endoscopy During gastroscopy, two biopsy specimens were taken from the gastric antrum $2 cm from the pylorus, the greater curve of the gastric body, within 2 cm from esophagogastric junction, and from the distal esophagus $3 cm above the squamocolumnar junction (Z-line). The esophagogastric junctional biopsy specimens were obtained as follows. The proximal margins of the gastric mucosal folds were identified as the junction of the muscular wall of the esophagus and the stomach (7, 8). The normal squamocolumnar mucosal junction or Z-line is normally located within 2 cm of the proximal edge of the gastric folds (8). During endoscopy, the measurements were made by the endoscope with the incisor teeth as a reference point. Esophagogastric junctional biopsies were obtained within 2 cm of the proximal edge of the gastric folds under direct vision with the gastroscope in the antegrade position. In this presentation junctional SCE denotes the SCE detected histologically in biopsy specimens obtained within 2 cm of the proximal edge of the gastric folds. If gastric-type mucosa extended circumferentially or as tongues $2 cm proximally from the upper margin of the gastric folds, separate biopsy specimens were obtained from this mucosa to detect traditional-appearing BE (7, 8). The patients with SCE detected $2 cm above the proximal end of the gastric folds were excluded from the present analysis. The subjects with one or more nonconfluent or confluent erosions were classified to have erosive esophagitis (SavaryMiller grades II–IV [9]). Histology Formalin-fixed biopsies were embedded in paraffin and cut, and the tissue sections were stained with hematoxylin and eosin, alcian blue pH 2.5/periodic acid Schiff, and modified Giemsa methods. The presence of gastritis, the activity of gastritis, gastric gland atrophy, intestinal metaplasia, and H.
Figure 1. Specialized columnar epithelium is characterized by goblet cells with acidic mucin staining by various tones of blue and red when stained by alcian blue-periodic acid Schiff, pH 2.5. Also adjacent columnar-appearing cells contain acidic mucins.
pylori infection were classified according to the Sydney System (10). Modified Giemsa stain was used for detecting H. pylori infection. The presence of SCE, chronic inflammation (carditis), and H. pylori infection were examined in biopsy specimens obtained from the esophagogastric junction. Carditis was defined as the infiltration of the lamina propria by lymphocytes and plasma cells (11). Specialized columnar epithelium was defined as the intestinal metaplasia of immature type (types 2 and 3; Fig. 1). Typically, SCE is an epithelium showing goblet cells intermixed with mucus-secreting foveolar-type epithelium. Dysplasia, when present and definite, was graded as low-grade or high-grade (1). Histological esophagitis was defined as papillae extending into the upper third of the esophageal epithelium with or without infiltration of the epithelium by inflammatory cells (11). Statistics x2 test and the Mann-Whitney U test were used for the comparison of the junctional SCE and non-SCE groups. p values , 0.05 were considered to be significant. Multiple logistical regression analysis was used to assess the rela-
AJG – April, 1999
Junctional Specialized Columnar Epithelium
915
Table 1. Comparison Between Patients With (SCE1) and Without (SCE2) Junctional SCE Junctional SCE1, n 5 110 (100%)
Junctional SCE2, n 5 1009 (100%)
p Value
65.6 6 1.2 1:1.1 14% 22% 22% 92% 23% 66% 28% 40% 60% 11% 46%
56.7 6 0.5 1:1.2 14% 13% 26% 73% 19% 54% 15% 34% 48% 8% 36%
, 0.001 NS NS 0.01 NS , 0.001 NS 0.01 0.001 0.2 0.01 NS 0.04
Age (SEM)* Male:female GERD symptoms† Endoscopic erosive esophagitis Histological esophagitis Cardia gastritis Cardia H. pylori infection Antral gastritis Antral intestinal metaplasia Antral H. pylori infection Corpus gastritis Corpus intestinal metaplasia Corpus H. pylori infection * Standard error of mean. † Gastroesophageal reflux symptoms (heartburn and regurgitation). NS 5 not significant.
tionships between junctional SCE and factors with a significant difference of prevalence when the junctional SCE and non-SCE groups were compared by univariate analysis. For multivariate analysis age was grouped by decades to a noncontinuous variable. Odds ratios (OR) with 95% confidence intervals (95% CI) are reported for variables included into the multivariate analysis. Ethics This study was approved by the Ethics Committee of Jyva¨skyla¨ Central Hospital.
RESULTS The demographic, endoscopic, and histological findings of the study population are presented in Table 1. In all, 110 (10%) of the 1119 patients had junctional SCE. The agespecific junctional SCE prevalence increased with age (Fig. 2). Although junctional SCE was strongly associated with chronic gastritis, it was observed also in patients with normal gastric histology (Table 2, Fig. 2). The mean age of the SCE patents with normal gastric histology was 59.6 yr (95% CI, 55.5– 63.8 yr) and that of the SCE patients with chronic
Figure 2. The prevalences of junctional SCE in different age groups. White columns, patients with normal gastric histology (G2) and gray columns, those with chronic gastritis (G1). Black columns, total SCE prevalences in the respective age groups. The correlation between junctional SCE and age was 0.16 (p , 0.001). Junctional SCE was not observed in patients ,30 yr of age.
916
Voutilainen et al.
AJG – Vol. 94, No. 4, 1999
Figure 3. Prevalences of junctional SCE in men and women in different age groups. The difference in the gender-rate of junctional SCE prevalence was not significant in any age group.
gastritis was 68.5 yr (95% CI, 65.8 –71.4 yr; p 5 0.001). Junctional SCE was detected in 10% of men (N 5 53) and in 9% of women (N 5 57) (NS, not significant). The junctional SCE prevalence increased with age in both genders without significant gender-related difference in any age group (Fig. 3). The male:female ratio of the SCE patients was 1:1.1. The mean age of the female SCE patients was 67.3 yr (95% CI, 63.8 –70.9 yr) and that of the male SCE patients was 64.1 yr (95% CI, 59.7– 68.5 yr; NS). None of the SCE patients had epithelial dysplasia in the junctional biopsy specimens. Histologically normal gastric mucosa, in other words, normal appearance in random biopsies from antrum and corpus, occurred in 36 (33%) of 110 SCE patients. The SCE prevalences in different types of gastritis are presented in Table 2. In univariate analysis, a significant association between junctional SCE and antral predominant gastritis was detected (Table 2). No significant association was found, however, between junctional SCE and corpus intestinal metaplasia. Univariate analysis revealed a significant association between junctional SCE and endoscopic erosive
esophagitis, but not between SCE and histological esophagitis or heartburn and regurgitation (Table 1). Multivariate logistical regression analysis revealed that carditis (OR, 3.1; 95% CI, 1.4 – 68), endoscopic erosive esophagitis (OR, 1.8; 95% CI, 1.1–3.1), and age (OR, 1.4 per decade; 95% CI, 1.2–1.6) were independent risk factors for junctional SCE. The following factors were not independent risk factors for junctional SCE: chronic antral (OR, 1.0; 95% CI, 0.5–2.1) and corpus (OR, 0.8; 95% CI, 0.4 – 1.8) gastritis, corpus H. pylori infection (OR, 1.4; 95% CI, 0.7–2.8), and antral intestinal metaplasia (OR, 1.2; 95% CI, 0.7–2.1).
DISCUSSION This study revealed that in Finland in an unselected series of dyspeptic patients the junctional SCE prevalence was 10%. This prevalence might have been found to be higher if more than two biopsies had been taken from the esophagogastric junction. Johnston et al., who also took two biopsies from the esophagogastric junction, reported a 9.4% prevalence of
Table 2. Prevalences of Junctional SCE in Different Types of Gastritis Gastric Histology
Junctional SCE1
p Value
Normal histology (n 5 478) Chronic gastritis without IM* (n 5 439) Chronic gastritis with antral IM only (n 5 126) Chronic gastritis with corpus IM only (n 5 46) Chronic gastritis with antral and corpus IM (n 5 30)
36 (8%) 42 (10%) 25 (20%) 5 (11%) 2 (7%)
NS† , 0.001 NS NS
* IM 5 complete intestinal metaplasia. † Comparisons were made between the patients with normal gastric histology and the subtypes of chronic gastritis.
AJG – April, 1999
SCE (4). In other reports, based on varying biopsy protocols, the prevalence of SCE has been reported to vary from 5.3% (12) to 18% (3), and even as high as 36% (13). All these studies indicate that esophagogastric junctional SCE is a common lesion in Western populations. In the present study the junctional SCE prevalence increased with age, indicating that it could be an acquired lesion. Junctional SCE does not seem to result from gastritis or H. pylori infection because one-third of the SCE cases occurred in patients with histologically normal stomach, without evidence of H. pylori infection. In addition, no type of gastritis proved to be an independent risk factor for junctional SCE in multivariate analysis. Our finding that antral-predominant gastritis was more prevalent in the junctional SCE than in non-SCE patients is in accordance with the report of Weston et al. (14). Being the major causative factor for this type of gastritis, H. pylori may have an indirect role in the development of junctional SCE in some subjects. The possibility that junctional SCE is a simple extension of gastritis and gastric intestinal metaplasia from the gastric body to cardia is unlikely, as no association was found between junctional SCE and intestinal metaplasia in the corpus in this study. In addition, our finding that one-third of junctional SCE cases were detected in subjects with histologically normal antrum and corpus indicates that junctional SCE is an acquired local injury. So far, no firm correlation has been established between junctional SCE and GERD (11). Contrary to some previous reports (3, 4, 15), in the present study junctional SCE was found to associate with erosive esophagitis, suggesting that SCE may be a result of GERD. In this series, no association was detected between junctional SCE and histological esophagitis. Earlier studies of the clinical relevance of junctional SCE to histological esophagitis have yielded conflicting results (3, 4, 15). The lack of association between junctional SCE and histological esophagitis in this study is explained by the fact that minor histological changes in the squamous epithelium are found also in apparently normal individuals (11). In addition, histological esophagitis defined by the Ismail-Beigi criteria has been shown not to be completely relevant to GERD (16). Esophageal SCE is the histological hallmark of Barrett’s esophagus (5, 6). The aforementioned association between junctional SCE and erosive esophagitis means that there may be a subgroup of junctional SCE patients in whom junctional SCE could progress to classic Barrett’s esophagus, even though junctional SCE per se is probably not an unequivoval prelesion for Barrett’s esophagus. That the junctional SCE would straightforwardly extend to classic Barrett’s esophagus is unlikely, because the demographic characteristics of junctional SCE in this study were very different from those previously reported for classic Barrett’s esophagus, as well as cardia and esophageal adenocarcinoma. Barrett’s esophagus, and cardia and esophageal ade-
Junctional Specialized Columnar Epithelium
917
nocarcinoma, are highly male predominant (17–19). In the present series, unlike in some previous reports (3, 4), the male-to-female ratio of junctional SCE patients was practically 1. Hirota et al. have also reported an equal distribution of junctional SCE between men and women (12). These findings suggest that some additional factors are probably needed before junctional SCE can progress to classic Barrett’s esophagus. In summary, the present study showed that junctional SCE was a lesion related to age, cardia inflammation, and erosive esophagitis, and therefore we suggest that it is an acquired lesion. Junctional SCE was not related to corpus intestinal metaplasia and could even be detected in patients with normal gastric histology, a finding that suggests that junctional SCE cannot be regarded as an extension of gastritis from the gastric body to the gastric cardia. The clinical significance of junctional SCE in relation to Barrett’s esophagus and cardia cancer remains unclear, and routine biopsies of the esophagogastric junction are not recommended. Reprint requests and correspondence: Markku Voutilainen, M.D., Jyva¨skyla¨ Central Hospital, Keskussairaalantie 19, FIN40620 Jyva¨skyla¨, Finland. Received June 15, 1998; accepted Dec. 19, 1998.
REFERENCES 1. Haggitt RC. Barrett’s esophagus, dysplasia, and adenocarcinoma. Hum Pathol 1994;25:982–93. 2. Weinstein WM, Ippoliti AF. The diagnosis of Barrett’s esophagus: Goblets, goblets, goblets. Gastrointest Endosc 1996;44:91–5. 3. Spechler SJ, Zeroogian JM, Antonioli DA, et al. Prevalence of metaplasia at the gastro-oesophageal junction. Lancet 1994; 344:1533– 6. 4. Johnston MH, Hammond AS, Laskin WL, et al. The prevalence and clinical characteristics of short segments of specialized intestinal metaplasia in the distal esophagus on routine endoscopy. Am J Gastroenterol 1996;91:1507–11. 5. Gottfried MR, McClave SA, Boyce HW. Incomplete intestinal metaplasia in the diagnosis of columnar lined esophagus (Barrett’s esophagus). Am J Clin Pathol 1989;92:741– 6. 6. Jauregui HO, Davessar K, Hale JH, et al. Mucin histochemistry of intestinal metaplasia in Barrett’s esophagus. Mod Pathol 1988;1:188 –92. 7. McClave SA, Boyce Jr HW, Gottfried MR. Early diagnosis of columnar-lined esophagus: A new endoscopic criterion. Gastrointest Endosc 1987;33:413– 6. 8. Tytgat GNJ, Hameeteman W, Onstenk R, et al. The spectrum of columnar-lined esophagus—Barrett’s esophagus. Endoscopy 1989;21:177– 85. 9. Savary M, Miller G, eds. The esophagus. Handbook and atlas of endoscopy. Solothurn, Switzerland: Gassman Verlag AG, 1978:135–9. 10. Price AB. The Sydney System: Histological division. J Gastroenterol Hepatol 1991;6:209 –22. 11. Riddell RH. The biopsy diagnosis of gastroesophageal reflux disease, “carditis,” and Barrett’s esophagus, and sequelae of therapy. Am J Surg Pathol 1996;20(suppl):S31–51. 12. Hirota WK, Loughney TM, Lazas DJ, et al. The prevalence and demographics of specialized intestinal metaplasia at the
918
Voutilainen et al.
esophagogastric junction in a cohort of 889 prospectively studied patients. Gastroenterology 1997;112(suppl):A149. 13. Nandurkar S, Talley NJ, Martin CJ, et al. Short segment Barrett’s oesophagus: Prevalence, diagnosis and associations. Gut 1997;40:710 –5. 14. Weston AP, Krmpotich P, Makdisi WF, et al. Short segment Barrett’s esophagus: Clinical and histological features, associated endoscopic findings, and association with gastric intestinal metaplasia. Am J Gastroenterol 1996;91:5:981– 6. 15. Trudgill NJ, Suvarna SK, Kapur KC, et al. Intestinal metaplasia at the squamocolumnar junction in patients attending for diagnostic gastroscopy. Gut 1997;41:585–9.
AJG – Vol. 94, No. 4, 1999
16. Schindlbeck NE, Wiebecke B, Klauser AG, et al. Diagnostic value of histology in non-erosive gastro-oesophageal reflux disease. Gut 1996;39:151– 4. 17. Blot WJ, DeVesa SS, Kneller RW, et al. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA 1991;265:1287–9. 18. Blot WJ, DeVesa SS, Fraumeni Jr JF. Continuing climb in rates of esophageal adenocarcinoma: An update. JAMA 1993; 270:1320. 19. Pera M, Cameron AJ, Trastek VF, et al. Increasing incidence of adenocarcinoma of the esophagus and esophagogastric junction. Gastroenterology 1993;104:510 –3.
Vol. 94, No. 4, 1999 ISSN 0002-9270/99/$20.00 PII S0002-9270(99)00041-6
Specialized Columnar Epithelium of the Esophagogastric Junction: Prevalence and Associations Markku Voutilainen, M.D., Martti Fa¨rkkila¨, M.D., Matti Juhola, M.D., Kyo¨sti Nuorva, M.D., Kari Mauranen, M.Sc., Timo Ma¨ntynen, M.D., Ilkka Kunnamo, M.D., Jukka-Pekka Mecklin, M.D., Pentti Sipponen, M.D., and The Central Finland Endoscopy Study Group* Departments of Internal Medicine, Pathology, and Surgery, Jyva¨skyla¨ Central Hospital, Jyva¨skyla¨; Division of Gastroenterology, Department of Internal Medicine, Helsinki University Hospital, Helsinki; Department of Health Policy and Management, Kuopio University, Kuopio; Karstula Health Centre, Karstula; Department of Pathology, Jorvi Hospital, Espoo, Finland
OBJECTIVES: In Barrett’s esophagus (BE) normal squamous esophageal epithelium is replaced by specialized columnar epithelium (SCE). BE is related to gastroesophageal reflux disease (GERD) and is a risk factor for esophageal adenocarcinoma. SCE is detected also at normal-appearing esophagogastric junction without BE (junctional SCE). The relationships between junctional SCE, GERD, and cardia adenocarcinoma are obscure and controversial. The aims of the present study were to investigate the prevalence and demographics of junctional SCE and to compare these figures with those reported for BE, and esophageal and cardia adenocarcinoma. A further aim was to examine the association between junctional SCE and GERD, Helicobacter pylori infection, and gastritis. METHODS: One thousand one hundred-nineteen consecutive dyspeptic patients underwent gastroscopy and were enrolled into the study. RESULTS: Junctional SCE was detected in 110 patients (10%). The age-specific prevalence of junctional SCE increased with age. The male:female ratio was 1:1.1. In multivariate analysis, junctional SCE was independently and positively related to endoscopic erosive esophagitis (odds ratio [OR], 1.8; 95% confidence interval [CI], 1.1–3.1), cardia inflammation (carditis) (OR, 3.1; 95% CI, 1.4 – 6.8), and age (OR, 1.4 per decade; 95% CI, 1.2–1.6), but not to corpus H. pylori infection (OR, 1.4; 95% CI, 0.7–2.8), antral (OR, 1.0; 95% CI, 0.5–2.1) or corpus (OR, 0.8; 95% CI, 0.4 –1.8) gastritis, or intestinal metaplasia of the antral mucosa in stomach (OR, 1.2; 95% CI, 0.7–2.1). In univariate analysis, junctional SCE was, however, significantly more common in patients with antral-predominant atrophic gas-
* The following investigators are members of The Central Finland Endoscopy Study Group: Liisa Ahlskog-Muraja, Teuvo Antikainen, Sirpa Antila, Jorma Anttinen, Matti Hallikas, Pekka Hannonen, Kari-Pekka Ha¨ma¨la¨inen, Heikki Janhonen, Matti Kairaluoma, Kerkko Karjalainen, Pekka Kauranen, Heikki Korhonen, Jari Korhonen, Ritva Koskela, Raimo Krees, Vesa Ka¨rja¨, Pa¨ivi Laaksonen, Matti Laukkanen, Reino Liisanantti, Marja Lohman, Antero Palmu, Ulla Palmu, Matti Pellinen, Pertti Sa¨rkka¨, Harri Sela¨nne, Tuula Tervo, Marianne Udd, Jukka Viinikka.
tritis (20%), compared with those with normal gastric histology (8%, p , 0.001). CONCLUSIONS: Junctional SCE is age related and may therefore be an acquired lesion. It is associated with cardia inflammation and endoscopic erosive esophagitis, but not with H. pylori infection or gastric intestinal metaplasia. Unlike BE and cardia cancer, junctional SCE occurs with similar frequency in men and women. (Am J Gastroenterol 1999;94:913–918. © 1999 by Am. Coll. of Gastroenterology)
INTRODUCTION The diagnosis of Barrett’s esophagus (BE) requires the detection of the replacement of normal esophageal stratified squamous epithelium by specialized columnar epithelium (SCE) without any specification of the length of the metaplastic segment (1, 2). BE is an acquired condition related to gastroesophageal reflux disease (GERD) and harbors an increased risk of esophageal adenocarcinoma (1, 3). In some patients SCE is detected at the normal-appearing gastroesophageal (GE) junction, without BE (3, 4). The biological significance of the latter SCE type is unknown, nor is it known whether SCE at the normal-appearing GE junction is the precursor lesion for BE and cardia adenocarcinoma. Specialized columnar epithelium is an incomplete form of intestinal metaplasia characterized by goblet cell metaplasia and histochemical evidence of acidic sialomucins and sulphomucins in goblet cells and adjacent gastric-appearing columnar cells (5). It is easily and unequivocally detected in gastric biopsies with such simple histochemical stains as alcian-blue pH 2.5/periodic acid-Schiff or high-iron diamine (5, 6). The first aim of this study was to investigate the prevalence and demography of the SCE lesion of the esophagogastric junction in consecutive outpatients referred to endoscopy for dyspeptic symptoms and to compare the demographics of junctional SCE with those reported previ-
914
Voutilainen et al.
AJG – Vol. 94, No. 4, 1999
ously for BE and cardia cancer. The second aim was to investigate factors, such as gastritis and gastroesophageal reflux disease, that may be associated with junctional SCE.
MATERIALS AND METHODS Patients The study population consisted of the 1698 dyspeptic patients sent for gastroscopy between September 1 and December 31, 1996 in the Jyva¨skyla¨ Central Hospital referral area, the population of which is approximately 260,000 inhabitants. Adequate esophagogastric junctional biopsy specimens were obtained from 1184 (69%) patients. The patients with long- or short-segment Barrett’s esophagus or SCE detected in the esophagus without any specification of the length of the metaplastic segment, previous H. pylori eradication, and gastric surgery were excluded. Thus the final study population consisted of 1119 patients (516 men and 603 women). Endoscopy During gastroscopy, two biopsy specimens were taken from the gastric antrum $2 cm from the pylorus, the greater curve of the gastric body, within 2 cm from esophagogastric junction, and from the distal esophagus $3 cm above the squamocolumnar junction (Z-line). The esophagogastric junctional biopsy specimens were obtained as follows. The proximal margins of the gastric mucosal folds were identified as the junction of the muscular wall of the esophagus and the stomach (7, 8). The normal squamocolumnar mucosal junction or Z-line is normally located within 2 cm of the proximal edge of the gastric folds (8). During endoscopy, the measurements were made by the endoscope with the incisor teeth as a reference point. Esophagogastric junctional biopsies were obtained within 2 cm of the proximal edge of the gastric folds under direct vision with the gastroscope in the antegrade position. In this presentation junctional SCE denotes the SCE detected histologically in biopsy specimens obtained within 2 cm of the proximal edge of the gastric folds. If gastric-type mucosa extended circumferentially or as tongues $2 cm proximally from the upper margin of the gastric folds, separate biopsy specimens were obtained from this mucosa to detect traditional-appearing BE (7, 8). The patients with SCE detected $2 cm above the proximal end of the gastric folds were excluded from the present analysis. The subjects with one or more nonconfluent or confluent erosions were classified to have erosive esophagitis (SavaryMiller grades II–IV [9]). Histology Formalin-fixed biopsies were embedded in paraffin and cut, and the tissue sections were stained with hematoxylin and eosin, alcian blue pH 2.5/periodic acid Schiff, and modified Giemsa methods. The presence of gastritis, the activity of gastritis, gastric gland atrophy, intestinal metaplasia, and H.
Figure 1. Specialized columnar epithelium is characterized by goblet cells with acidic mucin staining by various tones of blue and red when stained by alcian blue-periodic acid Schiff, pH 2.5. Also adjacent columnar-appearing cells contain acidic mucins.
pylori infection were classified according to the Sydney System (10). Modified Giemsa stain was used for detecting H. pylori infection. The presence of SCE, chronic inflammation (carditis), and H. pylori infection were examined in biopsy specimens obtained from the esophagogastric junction. Carditis was defined as the infiltration of the lamina propria by lymphocytes and plasma cells (11). Specialized columnar epithelium was defined as the intestinal metaplasia of immature type (types 2 and 3; Fig. 1). Typically, SCE is an epithelium showing goblet cells intermixed with mucus-secreting foveolar-type epithelium. Dysplasia, when present and definite, was graded as low-grade or high-grade (1). Histological esophagitis was defined as papillae extending into the upper third of the esophageal epithelium with or without infiltration of the epithelium by inflammatory cells (11). Statistics x2 test and the Mann-Whitney U test were used for the comparison of the junctional SCE and non-SCE groups. p values , 0.05 were considered to be significant. Multiple logistical regression analysis was used to assess the rela-
AJG – April, 1999
Junctional Specialized Columnar Epithelium
915
Table 1. Comparison Between Patients With (SCE1) and Without (SCE2) Junctional SCE Junctional SCE1, n 5 110 (100%)
Junctional SCE2, n 5 1009 (100%)
p Value
65.6 6 1.2 1:1.1 14% 22% 22% 92% 23% 66% 28% 40% 60% 11% 46%
56.7 6 0.5 1:1.2 14% 13% 26% 73% 19% 54% 15% 34% 48% 8% 36%
, 0.001 NS NS 0.01 NS , 0.001 NS 0.01 0.001 0.2 0.01 NS 0.04
Age (SEM)* Male:female GERD symptoms† Endoscopic erosive esophagitis Histological esophagitis Cardia gastritis Cardia H. pylori infection Antral gastritis Antral intestinal metaplasia Antral H. pylori infection Corpus gastritis Corpus intestinal metaplasia Corpus H. pylori infection * Standard error of mean. † Gastroesophageal reflux symptoms (heartburn and regurgitation). NS 5 not significant.
tionships between junctional SCE and factors with a significant difference of prevalence when the junctional SCE and non-SCE groups were compared by univariate analysis. For multivariate analysis age was grouped by decades to a noncontinuous variable. Odds ratios (OR) with 95% confidence intervals (95% CI) are reported for variables included into the multivariate analysis. Ethics This study was approved by the Ethics Committee of Jyva¨skyla¨ Central Hospital.
RESULTS The demographic, endoscopic, and histological findings of the study population are presented in Table 1. In all, 110 (10%) of the 1119 patients had junctional SCE. The agespecific junctional SCE prevalence increased with age (Fig. 2). Although junctional SCE was strongly associated with chronic gastritis, it was observed also in patients with normal gastric histology (Table 2, Fig. 2). The mean age of the SCE patents with normal gastric histology was 59.6 yr (95% CI, 55.5– 63.8 yr) and that of the SCE patients with chronic
Figure 2. The prevalences of junctional SCE in different age groups. White columns, patients with normal gastric histology (G2) and gray columns, those with chronic gastritis (G1). Black columns, total SCE prevalences in the respective age groups. The correlation between junctional SCE and age was 0.16 (p , 0.001). Junctional SCE was not observed in patients ,30 yr of age.
916
Voutilainen et al.
AJG – Vol. 94, No. 4, 1999
Figure 3. Prevalences of junctional SCE in men and women in different age groups. The difference in the gender-rate of junctional SCE prevalence was not significant in any age group.
gastritis was 68.5 yr (95% CI, 65.8 –71.4 yr; p 5 0.001). Junctional SCE was detected in 10% of men (N 5 53) and in 9% of women (N 5 57) (NS, not significant). The junctional SCE prevalence increased with age in both genders without significant gender-related difference in any age group (Fig. 3). The male:female ratio of the SCE patients was 1:1.1. The mean age of the female SCE patients was 67.3 yr (95% CI, 63.8 –70.9 yr) and that of the male SCE patients was 64.1 yr (95% CI, 59.7– 68.5 yr; NS). None of the SCE patients had epithelial dysplasia in the junctional biopsy specimens. Histologically normal gastric mucosa, in other words, normal appearance in random biopsies from antrum and corpus, occurred in 36 (33%) of 110 SCE patients. The SCE prevalences in different types of gastritis are presented in Table 2. In univariate analysis, a significant association between junctional SCE and antral predominant gastritis was detected (Table 2). No significant association was found, however, between junctional SCE and corpus intestinal metaplasia. Univariate analysis revealed a significant association between junctional SCE and endoscopic erosive
esophagitis, but not between SCE and histological esophagitis or heartburn and regurgitation (Table 1). Multivariate logistical regression analysis revealed that carditis (OR, 3.1; 95% CI, 1.4 – 68), endoscopic erosive esophagitis (OR, 1.8; 95% CI, 1.1–3.1), and age (OR, 1.4 per decade; 95% CI, 1.2–1.6) were independent risk factors for junctional SCE. The following factors were not independent risk factors for junctional SCE: chronic antral (OR, 1.0; 95% CI, 0.5–2.1) and corpus (OR, 0.8; 95% CI, 0.4 – 1.8) gastritis, corpus H. pylori infection (OR, 1.4; 95% CI, 0.7–2.8), and antral intestinal metaplasia (OR, 1.2; 95% CI, 0.7–2.1).
DISCUSSION This study revealed that in Finland in an unselected series of dyspeptic patients the junctional SCE prevalence was 10%. This prevalence might have been found to be higher if more than two biopsies had been taken from the esophagogastric junction. Johnston et al., who also took two biopsies from the esophagogastric junction, reported a 9.4% prevalence of
Table 2. Prevalences of Junctional SCE in Different Types of Gastritis Gastric Histology
Junctional SCE1
p Value
Normal histology (n 5 478) Chronic gastritis without IM* (n 5 439) Chronic gastritis with antral IM only (n 5 126) Chronic gastritis with corpus IM only (n 5 46) Chronic gastritis with antral and corpus IM (n 5 30)
36 (8%) 42 (10%) 25 (20%) 5 (11%) 2 (7%)
NS† , 0.001 NS NS
* IM 5 complete intestinal metaplasia. † Comparisons were made between the patients with normal gastric histology and the subtypes of chronic gastritis.
AJG – April, 1999
SCE (4). In other reports, based on varying biopsy protocols, the prevalence of SCE has been reported to vary from 5.3% (12) to 18% (3), and even as high as 36% (13). All these studies indicate that esophagogastric junctional SCE is a common lesion in Western populations. In the present study the junctional SCE prevalence increased with age, indicating that it could be an acquired lesion. Junctional SCE does not seem to result from gastritis or H. pylori infection because one-third of the SCE cases occurred in patients with histologically normal stomach, without evidence of H. pylori infection. In addition, no type of gastritis proved to be an independent risk factor for junctional SCE in multivariate analysis. Our finding that antral-predominant gastritis was more prevalent in the junctional SCE than in non-SCE patients is in accordance with the report of Weston et al. (14). Being the major causative factor for this type of gastritis, H. pylori may have an indirect role in the development of junctional SCE in some subjects. The possibility that junctional SCE is a simple extension of gastritis and gastric intestinal metaplasia from the gastric body to cardia is unlikely, as no association was found between junctional SCE and intestinal metaplasia in the corpus in this study. In addition, our finding that one-third of junctional SCE cases were detected in subjects with histologically normal antrum and corpus indicates that junctional SCE is an acquired local injury. So far, no firm correlation has been established between junctional SCE and GERD (11). Contrary to some previous reports (3, 4, 15), in the present study junctional SCE was found to associate with erosive esophagitis, suggesting that SCE may be a result of GERD. In this series, no association was detected between junctional SCE and histological esophagitis. Earlier studies of the clinical relevance of junctional SCE to histological esophagitis have yielded conflicting results (3, 4, 15). The lack of association between junctional SCE and histological esophagitis in this study is explained by the fact that minor histological changes in the squamous epithelium are found also in apparently normal individuals (11). In addition, histological esophagitis defined by the Ismail-Beigi criteria has been shown not to be completely relevant to GERD (16). Esophageal SCE is the histological hallmark of Barrett’s esophagus (5, 6). The aforementioned association between junctional SCE and erosive esophagitis means that there may be a subgroup of junctional SCE patients in whom junctional SCE could progress to classic Barrett’s esophagus, even though junctional SCE per se is probably not an unequivoval prelesion for Barrett’s esophagus. That the junctional SCE would straightforwardly extend to classic Barrett’s esophagus is unlikely, because the demographic characteristics of junctional SCE in this study were very different from those previously reported for classic Barrett’s esophagus, as well as cardia and esophageal adenocarcinoma. Barrett’s esophagus, and cardia and esophageal ade-
Junctional Specialized Columnar Epithelium
917
nocarcinoma, are highly male predominant (17–19). In the present series, unlike in some previous reports (3, 4), the male-to-female ratio of junctional SCE patients was practically 1. Hirota et al. have also reported an equal distribution of junctional SCE between men and women (12). These findings suggest that some additional factors are probably needed before junctional SCE can progress to classic Barrett’s esophagus. In summary, the present study showed that junctional SCE was a lesion related to age, cardia inflammation, and erosive esophagitis, and therefore we suggest that it is an acquired lesion. Junctional SCE was not related to corpus intestinal metaplasia and could even be detected in patients with normal gastric histology, a finding that suggests that junctional SCE cannot be regarded as an extension of gastritis from the gastric body to the gastric cardia. The clinical significance of junctional SCE in relation to Barrett’s esophagus and cardia cancer remains unclear, and routine biopsies of the esophagogastric junction are not recommended. Reprint requests and correspondence: Markku Voutilainen, M.D., Jyva¨skyla¨ Central Hospital, Keskussairaalantie 19, FIN40620 Jyva¨skyla¨, Finland. Received June 15, 1998; accepted Dec. 19, 1998.
REFERENCES 1. Haggitt RC. Barrett’s esophagus, dysplasia, and adenocarcinoma. Hum Pathol 1994;25:982–93. 2. Weinstein WM, Ippoliti AF. The diagnosis of Barrett’s esophagus: Goblets, goblets, goblets. Gastrointest Endosc 1996;44:91–5. 3. Spechler SJ, Zeroogian JM, Antonioli DA, et al. Prevalence of metaplasia at the gastro-oesophageal junction. Lancet 1994; 344:1533– 6. 4. Johnston MH, Hammond AS, Laskin WL, et al. The prevalence and clinical characteristics of short segments of specialized intestinal metaplasia in the distal esophagus on routine endoscopy. Am J Gastroenterol 1996;91:1507–11. 5. Gottfried MR, McClave SA, Boyce HW. Incomplete intestinal metaplasia in the diagnosis of columnar lined esophagus (Barrett’s esophagus). Am J Clin Pathol 1989;92:741– 6. 6. Jauregui HO, Davessar K, Hale JH, et al. Mucin histochemistry of intestinal metaplasia in Barrett’s esophagus. Mod Pathol 1988;1:188 –92. 7. McClave SA, Boyce Jr HW, Gottfried MR. Early diagnosis of columnar-lined esophagus: A new endoscopic criterion. Gastrointest Endosc 1987;33:413– 6. 8. Tytgat GNJ, Hameeteman W, Onstenk R, et al. The spectrum of columnar-lined esophagus—Barrett’s esophagus. Endoscopy 1989;21:177– 85. 9. Savary M, Miller G, eds. The esophagus. Handbook and atlas of endoscopy. Solothurn, Switzerland: Gassman Verlag AG, 1978:135–9. 10. Price AB. The Sydney System: Histological division. J Gastroenterol Hepatol 1991;6:209 –22. 11. Riddell RH. The biopsy diagnosis of gastroesophageal reflux disease, “carditis,” and Barrett’s esophagus, and sequelae of therapy. Am J Surg Pathol 1996;20(suppl):S31–51. 12. Hirota WK, Loughney TM, Lazas DJ, et al. The prevalence and demographics of specialized intestinal metaplasia at the
918
Voutilainen et al.
esophagogastric junction in a cohort of 889 prospectively studied patients. Gastroenterology 1997;112(suppl):A149. 13. Nandurkar S, Talley NJ, Martin CJ, et al. Short segment Barrett’s oesophagus: Prevalence, diagnosis and associations. Gut 1997;40:710 –5. 14. Weston AP, Krmpotich P, Makdisi WF, et al. Short segment Barrett’s esophagus: Clinical and histological features, associated endoscopic findings, and association with gastric intestinal metaplasia. Am J Gastroenterol 1996;91:5:981– 6. 15. Trudgill NJ, Suvarna SK, Kapur KC, et al. Intestinal metaplasia at the squamocolumnar junction in patients attending for diagnostic gastroscopy. Gut 1997;41:585–9.
AJG – Vol. 94, No. 4, 1999
16. Schindlbeck NE, Wiebecke B, Klauser AG, et al. Diagnostic value of histology in non-erosive gastro-oesophageal reflux disease. Gut 1996;39:151– 4. 17. Blot WJ, DeVesa SS, Kneller RW, et al. Rising incidence of adenocarcinoma of the esophagus and gastric cardia. JAMA 1991;265:1287–9. 18. Blot WJ, DeVesa SS, Fraumeni Jr JF. Continuing climb in rates of esophageal adenocarcinoma: An update. JAMA 1993; 270:1320. 19. Pera M, Cameron AJ, Trastek VF, et al. Increasing incidence of adenocarcinoma of the esophagus and esophagogastric junction. Gastroenterology 1993;104:510 –3.