Role of Helicobacter pylori Infection and Chronic Inflammation in Gastric Cancer in the Cardia

Original Article Japanese Journal of Clinical Oncology Advance Access published June 19, 2007 Jpn J Clin Oncol doi:10.1093/jjco/hym029 Role of Helicobacter pylori Infection and Chronic Inflammation in Gastric Cancer in the Cardia Yasuo Egi 1, Masanori Ito 1, Shinji Tanaka 2, Shinobu Imagawa 1, Shunsuke Takata 1, Masaharu Yoshihara 3, Ken Haruma 4 and Kazuaki Chayama 1 1 Department of Medicine and Molecular Science, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima, 2 Department of Endoscopy, Hiroshima University Hospital, Hiroshima, 3 Health Service Center, Hiroshima University, Higashi-Hiroshima, Hiroshima and 4 Division of Gastroenterology, Department of Internal Medicine, Kawasaki Medical School, Kurashiki, Okayama, Japan Received September 22, 2006; accepted January 11, 2007 Background: Helicobacter pylori-induced gastritis is an important factor for gastric carcinogenesis. However, it is still controversial whether it is also applicable for cardiac cancer development. Recently, we reported that H. pylori is an important factor for the induction of cardiac inflammation. We examined the status of H. pylori-induced gastritis in patients with cardiac cancer. Methods: Seventy-five Japanese patients (58 men; mean age, 64.2 years) with cardiac cancer were studied. Cardiac cancer was defined as that mainly located within 2 cm from the squamo-columnar junction (SCJ). Histological gastritis including the cardiac region was evaluated using the biopsy or surgically resected sections. Cardiac inflammation was evaluated at 1 cm distal from SCJ in lesser curvature. Sera were collected and several markers were evaluated. The status of H. pylori infection was evaluated by histology and serum antibodies. Expressions of cytokeratins were examined by immunohistochemical analysis. Results: Out of 75 patients with cardiac cancer, H. pylori was positive in 71 (95%) patients. The cardiac inflammation was examined in 30 patients (26 with H. pylori and four without H. pylori infection) and we found cardiac inflammation was present in all cases with H. pylori infection. Histologically, H. pylori-related gastritis was also found in the gastric corpus and antrum. Serological data were consistent with the presence of chronic atrophic gastritis. Intestinal metaplasia was found in 18 cases in the cardiac mucosa, and their cytokeratin 7/20 pattern was judged as a gastric pattern in all cases. Conclusion: H. pylori infection is closely associated with cardiac cancer. Key words: atrophic gastritis cardiac cancer Helicobacter pylori cytokeratin INTRODUCTION Chronic gastritis is mainly induced by Helicobacter pylori infection, and provides the basis for gastric carcinogenesis (1,2). Many experimental and epidemiological studies revealed that H. pylori plays an important role in gastric carcinogenesis, not only of intestinal type gastric cancer, but For reprints and all correspondence: Masanori Ito, Department of Medicine and Molecular Science, Graduate School of Biomedical Sciences, Hiroshima University, 1-2-3 Kasumi Minami-ku, Hiroshima 734-8551, Japan. E-mail: maito@hiroshima-u.ac.jp also of diffuse type (3,4). Especially in non-cardiac cancer, an extremely close relationship was reported between H. pylori infection and gastric carcinogenesis (5,6). Control of H. pylori infection is now a standard medical strategy to reduce the prevalence of gastric cancer especially in Japan where gastric cancer is still very common (7). However, in cardiac cancer, the implication of H. pylori infection in its occurrence still remains controversial. A meta-analysis of the previous epidemiological studies revealed a negative association between the prevalence of H. pylori infection and of cardiac cancer (5). The prevalence # 2007 Foundation for Promotion of Cancer Research

Page 2 of 6 Chronic gastritis and cardiac carcinogenesis of total gastric cancer is expected to be reduced in the near future by the reduced prevalence of H. pylori infection (8,9). However, previous reports have indicated that diseases, such as gastro-esophageal reflux disease or cardiac cancer, may increase in the future (10). It will thus become an important question whether the prevalence of cardiac cancer will increase or not. The etiology of cardiac inflammation (so-called carditis) which could be linked to cardiac cancer remains uncertain. Recently, we examined the pathogenesis of inflammation in the cardiac area in Japanese patients and demonstrated the significance of H. pylori infection in cardiac inflammation (11). However, it is still unclear whether H. pylori-related gastritis is really present in patients with cardiac cancer and also whether H. pylori is a important factor for cardiac carcinogenesis or not. In addition, the criterion of the cardiac region is still uncertain, which is another important issue in this field. In the present study, we attempted to clarify the histological and serological background of gastric inflammation in patients with cardiac cancer. Based on these data, we discussed the implication of H. pylori infection in cardiac carcinogenesis in Japanese patients. PATIENTS AND METHODS SUBJECTS Seventy-five Japanese patients (58 men, 17 women; mean age, 64.2 years; range, 25 91 years) were enrolled in this study. In the present study, cardiac cancer was defined as that mainly located within 2 cm from the squamo-columnar junction (SCJ), as described by Siewert et al. (12). Patients treated by previous gasterectomy or the administration of non-steroidal anti-inflammatory drugs or proton pump inhibitors were excluded. Those having undergone successful eradication therapy for H. pylori were also excluded. Cardiac cancers were treated by surgical resection or endoscopic mucosal resection (EMR). Informed consent was obtained from all patients, and the ethics committee of Hiroshima University Hospital approved our protocol. HISTOLOGICAL EXAMINATION Gastritis was evaluated histologically using surgically resected specimens or biopsy specimens. We selected the biopsy or surgically resected sections of the gastric corpus and the antrum in the lesser curvature that were suitable for evaluation of gastritis histologically. The sections that were influenced by the gastric lesions were excluded from our study. The appearance of gastric inflammation was scored using the updated Sydney system (13). For the assessment of cardiac inflammation, we chose the biopsies or surgically resected sections from 1-cm distal to the SCJ on the lesser curvature, and histological gastritis was estimated in the same manner. In the cardiac biopsy, we classified specimens into two groups on the basis of the predominant gastric gland: fundic gland (FG)-dominant or simple mucinous gland-dominant (11). We excluded the section that was influenced by the invasion of tumor tissue in the gastric cardia. HELICOBACTER PYLORI INFECTION The presence of H. pylori was determined by histological examination and assessment of serum antibodies against H. pylori (E-plate, Eiken, Japan). In H. pylori-negative patients, we confirmed that both tests were negative and that there was an absence of histological gastritis in the gastric corpus and antrum. MEASUREMENT OF SERUM PEPSINOGEN, GASTRIN AND ANTI-PARIETAL CELL ANTIBODY Fasting sera were collected from 34 patients upon entry into the study. The sample was centrifuged immediately at 48C and stored at 208C until use. Serum concentrations of pepsinogens (PGs) and gastrin were determined by modified radioimmunoassay (14). The serum titer of anti-parietal cell antibody (APCA) was evaluated with the enzyme-linked immunosorbent assay as previously described (15). IMMUNOHISTOCHEMICAL ANALYSIS Four-micrometer sections of formalin-fixed paraffinembedded tissues were used for immunohistochemical staining. After deparaffinization and hydration, internal peroxidase was blocked by incubating with 0.3% H 2 O 2 in methanol for 15 min. After incubation with 5% skim milk/ PBS for 20 min, the sections were reacted with the primary antibody (diluted with PBS) for 2 h at room temperature. The primary antibodies used were anti-cytokeratin 7 polyclonal antibody (DAKO, Kyoto, Japan), and anti-cytokeratin 20 polyclonal antibody (DAKO, Kyoto, Japan) (16,17). We performed the immunostaining using an LSAB2 kit (DAKO, Kyoto, Japan). The staining pattern was classified into two patterns (gastric type and Barrett type), as illustrated in the previous report (18). STATISTICS Results were reported as the mean + standard error (SE). We assessed the significance of the difference by Mann Whitney U-test, t-test and x 2 -test with StatView software (SAS Institute Inc., Cary, NC). A P value of,0.05 was considered significant. RESULTS CLINICOPATHOLOGICAL FINDINGS IN PATIENTS WITH CARDIAC CANCER The clinical features of 75 patients with cardiac cancer are summarized in Table 1. Forty-four of the 75 (59%) were diagnosed as early gastric cancer, which limited within submucosal

Jpn J Clin Oncol Page 3 of 6 Table 1. Clinicopathological features of 75 cardiac cancer patients Age (mean) 64.2 + 1.5 Gender (male/female) 58/17 Histology (intestinal/diffuse) 52/23 Depth (early 1 /advanced) 44/31 Treatment (EMR/operation) 33/42 H. pylori (positive/negative) 71/4 EMR, endoscopic mucosal resection. 1 Cancer limited within submucosal layer. layer. Out of 44, 33 were diagnosed as mucosal cancer preoperatively and treated by endoscopic mucosal resection (EMR). We examined the status of H. pylori infection and 71 (95%) were regarded as H. pylori-positive. In four patients without H. pylori infection, three tumors were diagnosed as the intestinal type and one was as diffuse type cancer. RELATION BETWEEN H. PYLORI INFECTION AND CARDIAC INFLAMMATION We then examined the status of histological inflammation in 30 cardiac mucosal sections. In the histological evaluation of cardiac mucosa, we excluded the sections showing the involvement of benign or malignant lesions. The results, including the sub-classification by the H. pylori status, are shown in Table 2. In H. pylori-positive patients, we detected statistically higher degrees of lymphocyte infiltration in the cardiac mucosa than those in H. pylori-negative patients (P, 0.05). Neutrophil infiltration, atrophy and intestinal metaplasia in the cardiac region were only detected in H. pylori positive patients. Fundic gland (FG)-dominance featured histologically in 35% of cases in H. pylori positive group; whereas, all H. pylori-negative samples were FG-dominant. When we defined the histological gastritis as that with mild/no atrophy and lymphocyte infiltration without neutrophil infiltration or intestinal metaplasia (19), histological gastritis in the gastric cardia was present in all sections with H. pylori infection; whereas, none showed Table 2. Status of histological inflammation of cardiac mucosal sections histological inflammation in sections without H. pylori infection (P, 0.01). RELATION BETWEEN CARDIAC AND CORPUS/ANTRAL INFLAMMATION We examined the background gastritis in patients with cardiac cancer both histologically and serologically. As for histological examination, we could collect biopsy samples from the gastric corpus and antrum simultaneously only from 25 patients whose sections from the gastric corpus/antrum in the lesser curvature has a high quality for scoring gastritis. As demonstrated in Table 3, we could detect a higher degree of histological gastritis not only in the corpus, but also in the antrum. The grades of gastritis (atrophy, lymphocyte infiltration, neutrophil infiltration and intestinal metaplasia) were prominent in patients with H. pylori infection. Especially, we found the higher grades of intestinal metaplasia in the gastric corpus. As for serological data from 35 patients, the mean values of serum pepsinogen (PG) I, PG II, PG I/II and gastrin were 55.6 ng/ml, 21.9 ng/ml, 2.54 and 211.9 pg/ml, respectively. A low level of pepsinogen I (PG-I) and I/II ratio, and a high level of gastrin were found, and these suggest the presence of corpus atrophic gastritis. The serological status of the APCA level, which is considered to be an inducible factor for corpus atrophy, was not different between the non-neoplastic control groups (data not shown). EXPRESSION PATTERN OF CYTOKERATINS IN THE METAPLASTIC MUCOSA OF THE CARDIAC REGION We examined the expression pattern in sections with intestinal metaplasia in the cardiac mucosa in patients with cardiac cancer. In 18 sections, we detected intestinal metaplasia in the cardiac section (1 cm from SCJ in the lesser curvature) and these were subjected to immunohistochemical examinations. The cytokeratin (CK) expression pattern was subclassified into two types; the gastric type (absence of CK-7 staining and positive CK-20 staining) and the Barrett type (strong CK-7 staining and positive CK-20 staining in the superficial mucosa) (20). In the sections examined, all 18 sections that had H. pylori infection revealed the gastric type staining (Table 4, Figure 1). No sections showed the Barrett pattern. All gastric mucosa in the corpus revealed gastric H. pylori (n ¼ 26) H. pylori (n ¼ 4) P value Atrophy 1.96 + 0.13 * 0 *,0.001 Lymphocyte infiltration 1.62 + 0.11 * 0.67 + 0.33 * 0.01 Neutrophil infiltration 0.50 + 0.14 0 0.23 Intestinal metaplasia 1.19 + 0.21 0 0.07 Fundic gland dominancy 9/26 4/4 0.06 Cardiac inflammation a 26/26 * 0/4 *,0.001 a Defined as described in the text. *P, 0.01 Table 3. Histological gastritis of corpus/antral mucosa in patients with cardiac cancer Antrum (n ¼ 25) Corpus (n ¼ 25) Atrophy 2.04 + 0.21 1.80 + 0.17 Lymphocyte infiltration 1.44 + 0.14 1.44 + 0.14 Neutrophil infiltration 0.48 + 0.10 0.44 + 0.14 Intestinal metaplasia 2.04 + 0.25 1.36 + 0.22 Scores: mean + SE.

Page 4 of 6 Chronic gastritis and cardiac carcinogenesis Table 4. Expression of cytokeratin 7/20 in gastric mucosa type staining. In addition, we examined 23 sections from gastritis (non-cancer) patients and confirmed that more than 90% of the cardiac sections with intestinal metaplasia showed the gastric pattern (Table 4). DISCUSSION Cardiac cancer Non-cancer Cardia (n ¼ 18) (n ¼ 23) Barrett type 0 2 Gastric type 18 21 Corpus (n ¼ 8) (n ¼ 15) Barrett type 0 0 Gastric type 8 15 Although H. pylori is thought to be an important factor for gastric carcinogenesis, it is still controversial whether it holds true for cardiac cancer. Previous meta-analysis demonstrated a weak association between H. pylori infection and cardiac cancer (5). However, the major problem in this field is the uncertainty of definition of the term cardiac cancer. The most popular criterion was that designed by Siewert, which defines cardiac cancer as that located within 2 cm of the SCJ (12). Therefore, in the present study we defined cardiac cancer as that mainly located within 2 cm of the SCJ. This area is known as the cardiac area ; however, it is unclear whether true cardiac glands are present in this area. Recent studies have indicated that true cardiac glands are extremely limited, to less than 5 mm from the SCJ (20). We have also reported to find the intact oxyntic gland on the cardiac region (1 cm from ECJ) in sections without inflammatory change and simple mucinous glands could be detected only in cases with chronic inflammation with H. pylori infection in Japanese patients (11). Therefore, we suggest that these simple mucinous glands are not true cardiac glands but metaplastic mucosa from oxyntic glands. It is likely that cardiac cancer comprises two different types; one is the cancer from true cardiac mucosa (with cardiac gland) whereas the other is derived from the gastric corpus (metaplastic) epithelium with chronic inflammation induced by H. pylori infection. The proportion of these two may be different between countries or races. In the present study, we collected relatively large numbers of cardiac cancer and examined the background mucosa of these patients. We could demonstrate that most patients with cardiac cancer belonged to the latter group; that is, the cancer derived from metaplastic cardiac mucosa with H. pylori infection. Our histological and serological data demonstrated that these patients had pan-gastritis with H. pylori-related gastric inflammation from the cardia to corpus/antrum. From our point of view, the carcinogenesis of cardiac cancer may be tightly related to H. pylori infection and its related gastritis. Moreover, the major part of gastric cancer derived from this point examined (1 cm from Figure 1. Immunohistochemical images for cytokeratin expression; CK-7 (A,C) and CK-20 (B,D). Representative immunostaining pattern of the gastric pattern (A,B) and Barrett pattern (C,D) are demonstrated.

Jpn J Clin Oncol Page 5 of 6 ECJ) should not be considered as true cardiac cancer but proximal cancer derived from the fundic area. Intestinal metaplasia could be detected in half of all cardiac sections examined, the origin of which is still under debate. It may be derived from the cardiac region or from mataplastic change from the oxyntic gland like a pseudopyloric metaplasia. As described in our results, immunohistochemical analysis revealed that a major part of cardiac mucosas with intestinal metaplasia was of gastric type, not the Barrett pattern. In addition, these patients showed non-specific gastric inflammation in both the gastric corpus and antrum, compared with dyspepsia patients with H. pylori infection. This suggests that, in the majority of cardiac cancer cases, or at least in our subjects, the mataplastic mucosa is closely associated with H. pylori-related chronic gastritis. Recently, Yagi et al. also reported that the intestinal metaplasia in the cardiac area (5 mm from ECJ) is associated with H. pylori infection and its related gastritis in Japanese patients (21). However, the specificity of the cytokeratins were controversial and some recent studies have raised the question as to its specificity (22). The problem in our study is the limited examination point of the cardiac region. We only examined the histological gastritis in the lesser curvature, but cardiac cancer could be developed from any region of the cardiac area. We shall examine the difference in the implication of H. pylori in cardiac inflammation between cardiac mucosas from lesser and from greater curvature in the next step. Also, the extension of the true cardiac gland should be defined, especially in H. pylori-negative stomach. The prevalence of H. pylori infection is decreasing gradually in the Japanese population, which may lead to the reduced prevalence of gastric cancer. However, the true trend of the incidence of cardiac cancer is still uncertain. From our data, we should conclude that cardiac cancer (defined mainly as that located within 2 cm of the ECJ) will also decrease along with non-cardiac cancer in the future. In our case examined, only limited cases with cardiac cancer were considered to be H. pylori-independent and these may increase in the future (23). Further studies should focus on the analysis of these H. pylori-independent cardiac cancers, since it will provide useful information for effective cancer screening. Recent advances in magnifying endoscopic examination will provide useful information in this field (24). Taken together, our present results suggest that a major part of cardiac cancer is mainly induced by H. pylori infection in Japanese patients. The treatment of chronic inflammation may therefore lead to the control of this kind of gastric cancer. Acknowledgment We thank Ms Nao Kubota (Hiroshima University, Japan) for the excellent assistance in the production of the database used in this study. This study was supported by grant from Tsuchiya Research Foundation. Conflict of interest statement None declared. References 1. Correa P. Helicobacter pylori and gastric carcinogenesis. Am J Surg Path 1995;19(Suppl 1):S37 43. 2. Kawaguchi H, Haruma K, Komoto K, Yoshihara M, Sumii K, Kajiyama G. Helicobacter pylori infection is the major risk factor for atrophic gastritis. Am J Gastroenterol 1996;91:959 62. 3. Haruma K, Komoto K, Kamada T, Ito M, Kitadai, Yoshihara M, et al. Helicobacter pylori infection is a major risk factor for gastric carcinoma in young patients. Scand J Gastroenterol 2000;35:255 9. 4. Kohmoto K, Haruma K, Kamada T, Tanaka S, Yoshihara M, Sumii K, et al. Helicobacter pylori infection and gastric neoplasia: correlations with histological gastritis and tumor histology. Am J Gastroenterol 1998;93:1271 6. 5. Huang JQ, Sridhar S, Chen Y, Hunt RH. Meta-analysis of the relationship between Helicobacter pylori seropositivity and gastric cancer. Gastroenterology 1998;114:1169 79. 6. Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S, Yamakido M, et al. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med 2001;345:784 9. 7. Uemura N, Mukai T, Okamoto S, Yamaguchi S, Mashiba H, Taniyama K, et al. Effect of Helicobacter pylori eradication on subsequent development of cancer after endoscopic resection of early gastric cancer. Cancer Epidemiol Biomarkers Prev 1997;6:639 42. 8. Sasazuki S, Sasaki S, Tsugane S. Japan Public Health Center Study Group. Cigarette smoking, alcohol consumption and subsequent gastric cancer risk by subsite and histologic type. Int J Cancer 2002;101:560 6. 9. Ito M, Haruma K, Kamada T, Kitadai Y, Hidaka T, Tsuda T, et al. Reduction in the incidence of Helicobacter pylori-associated carcinoma in Japanese young adults. Oncol Rep 2001;8:633 6. 10. Fujimori T, Kawamata H, Ichikawa K, Ono Y, Okura Y, Tomita S, et al. Pathological issues of gastric and lower esophageal cancer: Helicobacter pylori infection and its eradication. J Gastroenterol 2002;37(Suppl 13):28 33. 11. Egi Y, Kim S, Ito M, Tanaka S, Yoshihara M, Haruma K, et al. Helicobacter pylori infection is the major risk factor for gastric inflammation in the cardia. Dig Dis Sci Apr 8; [Epub ahead of print]. 12. Siewert JR, Stein HJ. Carcinoma of the cardia: Carcinoma of the gastroesophageal junction classification, pathology and extent of resection. Dis Esophagus 1996;9:173 82. 13. Dixon MF, Genta RM, Yardley JH, Correa P and the participants in the International Workshop on the Histopathology of Gastritis, Houston 1994. Classification and grading of gastritis. The updated Sydney System. International Workshop on the Histopathology of Gastritis. Am J Surg Path 1996;20:1161 81. 14. Yoshihara M, Sumii K, Haruma K, Kiyohira K, Hattori N, Kitadai Y, et al. Correlation of ratio of serum pepsinogen I and II with prevalence of gastric cancer and adenoma in Japanese subjects. Am J Gastroenterol 1998;93:1090 6. 15. Mardh E, Mardh S, Mardh B, Borch K. Diagnosis of gastritis by means of a combination of serological analyses. Clin Chim Acta 2002;320:17 27. 16. Ramaekers F, van Niekerk C, Poels L, Schaafsma E, Huijsmans A, Robben H, et al. Use of monoclonal antibodies to keratin 7 in the differential diagnosis of adenocarcinomas. Am J Pathol 1990;136:641 55. 17. Moll R, Lowe A, Laufer J, Franke WW. Cytokeratin 20 in human carcinomas. A new histodiagnostic marker detected by monoclonal antibodies. Am J Pathol 1992;140:427 47. 18. Ormsby AH, Vaezi MF, Richter JE, Goldblum JR, Rice TW, Falk GW, et al. Cytokeratin immunoreactivity patterns in the diagnosis of short-segment Barrett s esophagus. Gastroenterology 2000;119:683 90. 19. Kim S, Haruma K, Ito M, Tanaka S, Yoshihara M, Chayama K. Magnifying gastroendoscopy for diagnosis of histologic gastritis in the gastric antrum. Dig Liver Dis 2004;36:286 91. 20. Glickman JN, Fox V, Antonioli DA, Wang HH, Odze RD. Morphology of the cardia and significance of carditis in pediatric patients. Am J Surg Pathol 2002;26:1032 9.

Page 6 of 6 Chronic gastritis and cardiac carcinogenesis 21. Yagi K, Nakamura A, Sekine A. Intestinal metaplasia of gastric cardia and carditis in Japanese patients with Helicobacter pylori infection. Digestion 2004;70:103 8. 22. Mohammed IA, Streutker CJ, Riddell RH. Utilization of cytokeratins 7 and 20 does not differentiate between Barrett s esophagus and gastric cardiac intestinal metaplasia. Mod Pathol 2002;15:611 6. 23. Yagi K, Nakamura A, Sekine A, Oyamatsu M, Tamiya Y, Watanabe H. An early cancer of the gastric cardia arising from carditis after long-term gastroesophageal reflux disease in the absence of Helicobacter pylori infection. J Gastroenterol Hepatol 2002;17:1236 8. 24. Endo T, Awakawa T, Takahashi H, Arimura Y, Itoh F, Yamashita K, et al. Classification of Barrett s epithelium by magnifying endoscopy. Gastrointest Endosc 2002;55:641 7.