Gastric Cancers Emerging After H. Pylori Eradication Arise Exclusively from Non-Acid-Secreting Areas

Tohoku J. Exp. Med., 2012, 226, 45-53 Gastric Cancer Emerging after Eradication 45

Gastric Cancers Emerging after H. pylori Eradication Arise Exclusively from Non-Acid-Secreting Areas

Katsunori Iijima,1 Yasuhiko Abe,1 Tomoyuki Koike,1 Kaname Uno,1 Hiroyuki Endo,1 Waku Hatta,1 Naoki Asano,1 Kiyotaka Asanuma,1 Akira Imatani1 and Tooru Shimosegawa1

1Division of Gastroenterology, Tohoku University Graduate School of Medicine, Sendai, Japan

Although Helicobacter pylori (H. pylori) eradication has some inhibitory effects on the subsequent development of gastric cancer, there are sporadic cases of gastric cancer even after successful eradication. The pathogenesis of gastric cancer emerging after H. pylori eradication remains to be clarified. In this study, employing Congo-red chromoendoscopy, which is capable of visualizing the acid-secreting fundic mucosa, we investigated the topographic relationship of the acid secretion pattern to the occurrence site of gastric cancers emerging after eradication. Fourteen consecutive patients who suffered from new gastric cancer after eradication, defined as lesions that were discovered at least 2 years after the eradication, were prospectively enrolled. Whether the neoplasias arose from acid-secreting or non- acid-secreting areas was evaluated with Congo-red chromoendoscopy. Biopsy specimens taken from the two areas were subjected to histologic evaluation and immunohistochemistry for Ki-67 and p53. The mean period from the eradication to the subsequent occurrence of gastric cancer was 74 (44) months. There were two cancer lesions in 5 cases, and thus there was a total 19 lesions from 14 cases. Congo-red chromoendoscopy revealed that all 19 lesions arose exclusively from non-acid-secreting areas. Histological examination revealed sustained hyperproliferation and accumulation of p53 protein was frequently detectable in non-acid-secreting areas. Genetic alteration such as p53 mutation seems to be already present in the residual non-acid-secreting areas after eradication, areas that could be the origin of gastric carcinogenesis after eradication. Identification of such high-risk areas should be a promising approach for estimating the individual cancer risk after eradication.

The vast majority of gastric cancer arises from H. is important to clarify the pathogenesis of persisting carci- pylori-infected stomach mucosa (Uemura et al. 2001). nogenesis after the primary carcinogen, H. pylori infection, Recent studies consistently reported that H. pylori eradica- has been removed. tion had some inhibitory effects on the subsequent develop- Many investigators, including our group, consistently ment of gastric cancer, at least in certain conditions (Take et reported that the suppressed gastric acid secretion in H. al. 2005; Takenaka et al. 2007; Fukase et al. 2008; Ogura et pylori-positive patients with corpus gastritis with or without al. 2008). However, it also has been revealed that there are atrophy recovered at least partially after eradication of the sporadic cases of gastric cancer even after successful eradi- infection (El-Omar et al. 1997; Tucci et al. 1998; Iijima et cation (Kamada et al. 2005; Yanaoka et al. 2009; de Vries et al. 2004). In this context, employing Congo-red chromoen- al. 2009). Hence, careful follow-up with endoscopic exam- doscopy, which is known to be capable of visualizing the ination for possible gastric cancer is necessary even after acid-secreting fundic area by means of a pH-dependent successful eradication of H. pylori (Kamada et al. 2005; de color reaction (Tatsuta et al. 1973), we found that the area Vries et al. 2009; Yanaoka et al. 2009). Using endoscopic of the acid-secreting mucosa in the fundus was promptly surveillance after eradication, the sites where new cancer expanded after eradication, especially in the greater curva- preferentially develops could be determined. Additionally, ture (Sekine et al. 2006). However, in the subsequent long- identifying the underlying mucosal milieu that promotes the term follow-up (mean periods of 62 months after eradica- development of gastric cancer after successful eradication tion) study using that technique, we also recognized that

Received October 25, 2011; revision accepted for publication December 8, 2011. doi: 10.1620/tjem.226.45 Correspondence: Katsunori Iijima, M.D., Division of Gastroenterology, Tohoku University Graduate School of Medicine, 1-1 Seiryo- machi, Aobaku, Sendai 980-8574, Japan. e-mail: [email protected]

45 46 K. Iijima et al. there were varying degrees of residual non-acid-secreting ing area discriminated by Congo-red chromoendoscopy. Finally, two (functionally irreversible) areas in the fundus in most cases, gastric biopsy specimens were taken in the fundus from the acid- indicating incomplete recovery in terms of the regional secreting and non-acid-secreting areas within 2 cm of the boundary acid-secreting capacity (Iijima et al. 2009a). Given that the between the two areas and were subjected to the following histologic non-acid-secreting areas were characterized not only by and immunohistochemical evaluation. residual inflammation, extensive gastric mucosal atrophy Histology and intestinal metaplasia, but also by sustained, high levels Using the Updated Sydney system (Misiewicz et al. 1990), the of epithelial proliferation, we hypothesized that such func- degrees of inflammation, activity, atrophy, and intestinal metaplasia tionally irreversible mucosa could reflect increased malig- were scored from 0 to 3 by two independent pathologists in a blinded nant potential where new cancer could develop after eradi- manner (YA and AI). cation (Iijima et al. 2009a). In this study, in order to test our hypothesis concerning Immunohistochemistry the preferential site of cancer after the eradication of H. Serial sections from paraffin-embedded biopsy specimens were pylori, we enrolled patients who actually suffered from gas- assessed immunohistochemically for Ki-67 and p53 (Yabuki et al. tric cancer long after the eradication of H. pylori infection. 1997). Cell proliferation in the gastric epithelium was determined Then, we evaluated whether the neoplasias occurring after immunohistochemically using the monoclonal antibody MIB-1 eradication had arisen from acid-secreting or non-acid- (Dako, Glostrup, Denmark) against the nuclear proliferation-associ- secreting areas identified by the Congo-red chromoendos- ated antigen Ki-67. Tumor suppressor p53 was determined immuno- copy and then compared the histological characteristics histochemically using DO-7 antibody (Novocastra, Newcastle upon between the respective areas in relation to the residual car- Tyne, U.K.). For immunostaining, the tissue sections were incubated cinogenic potential. with the Ki-67 and p53 antibodies at a dilution of 1:100 at room tem- perature for 1 hour. The Ki-67 and p53 labeling indices were deter- Material and Methods mined by counting the number of Ki-67-positive and p53-positive cells in at least three well-oriented gastric pits (more than 500 epithe- Patients lial cells) and expressing the mean percentage of the total number of From January 2005 to August 2010, 14 all consecutive patients cells counted in each gastric pit. who suffered from new gastric cancer after eradication of H. pylori at Tohoku University Hospital were prospectively enrolled in this study. Serum pepsinogen concentrations These patients had received eradication therapy from 1996 to 2008 at Fasting serum samples obtained from all participants were cen- the hospital, and, after confirming that the treatment was successful, trifuged immediately at 4°C and stored at −20°C until measurement. they returned for periodic endoscopic examination. In this trial, gas- Serum concentrations of PG I were measured using a CLEIA kit tric cancers after eradication were defined as lesions that were discov- (Lumipulse pepsionogen I, Fujirebio Inc, Tokyo, Japan) (Iijima et al. ered at least 2 years after the eradication. 13C-urea breath test con- 2009b). firmed that all 14 patients had remained free from re-infection at the time of cancer occurrence. Those with a history of gastric surgery, Statistics and those taking anti-secretary agents such as H2 blocker and proton Clinical parameters of the enrolled subjects were expressed as pump inhibitor, were excluded from the study subjects. The study mean (s.d.). Significant difference in histological scores, and Ki-67 was approved by Tohoku University School of Medicine Ethics and p53 labeling indices were determined by the Mann-Whitney test. Committee and each subject gave written informed consent. P values < 0.05 were considered to be statistically significant. Congo-red chromoendoscopy Congo-red chromoendoscopy was performed as we recently Results reported (Sekine et al. 2006; Iijima et al. 2009a). Briefly, after an The characteristic features of the 14 enrolled patients overnight fast, pentagastrin at a dose of 6 µg/kg (pentagastrin; Sigma, in this analysis are listed in Table 1. Ten of 14 patients St. Louis, Mo, USA) was administered to the subjects via an intra- were male and the mean age at the occurrence of gastric muscular injection prior to endoscopy. Immediately following the cancer after eradication was 71 (9) years. The indications routine endoscopic inspection, a solution of 0.3% Congo-red and for eradication therapy were endoscopic mucosal resection 0.2M sodium bicarbonate was sprayed over the entire surface of the for early gastric cancer in 10 cases (Fukase et al. 2008), stomach through a spray catheter. Five minutes after spraying, the gastric ulcer in 2 cases, duodenal ulcers in 1, and idiopathic acid-secreting fundic mucosa turned a blue-black color, whereas the thrombocytopenic purpura in 1 case (Kohda et al. 2002), mucosa without acid production remained red. The areas of fundic mucosa with color shifting from red to blue- and the mean period from the eradication to the subsequent black were semi-quantitatively evaluated in two separate portions, occurrence of gastric cancer was 74 (44) months. In 10 namely “lesser curvature” and “greater curvature” (Sekine et al. patients who had received endoscopic mucosal resection 2006). In the lesser curvature, the areas with color shift were classi- previously, since none of the new cancerous lesions fied into 4 types: “all”, “large”, “small”, and “none”, while, inthe emerged from the preceding resection scar, they were greater curvature, they were classified into five types: “all”, “large”, unlikely to have been residual cancer due to incomplete “intermediate”, “small”, and “none”. Then, we determined whether endoscopic procedure. The cancerous lesions were dupli- the neoplastic lesion arose from the acid-secreting or non-acid-secret- cated in 5 cases, thus there were 19 lesions from 14 cases Gastric Cancer Emerging after Eradication 47 site NAS NAS NAS NAS NAS NAS NAS NAS NAS Occurrence NAS for all NAS for both NAS for both NAS for both NAS for both NAS for both

26 95 3.6 3.8 11.5 11.5 24.8 30.8 38.6 33.4 49.6 73.3 35.5 PG I (28.9) Serum M M M M M M M M SM M:17 Depth SM:2, M &M M & M & M & SM & M of cancers cancers Ant-LC MB-LC LB-AW LB-AW UB-PW UB-PW UB-PW UB-PW UB-PW Ant-GC Ant-GC Ant-GC Ang-LC MB-PW MB-PW Ant-AW Ant-AW Ant-AW Ant-AW Ang-AW Ang-AW As above UB-LC & UB-PW & UB-PW Ang-LC & Location of Ang-AW & Ang-AW Ang-AW & Ang-AW type 0-IIc 0-IIc 0-IIc 0-IIc 0-IIa 0-IIc 0-IIc 0-IIc 0-IIa 6/2/11 6/2/11 0-IIa/b/c = Macroscopic 0-IIc & 0-IIa & 0-IIa & 0-IIc & 0-IIb 0-IIc & 0-IIb 5 5 8 8 5 6 of 16 10 12 10 (6) (mm) 9 & 6 cancer 13 & 5 20 &10 10 & 20 & 10 Diameter of cancers Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Intestinal Histology Intestinal: all 1 2 1 1 2 2 1 1 2 1 2 1 1 1 of 19 lesions cancers Number 26 28 62 58 91 36 69 26 85 74 36 96 105 150 162 (44) (M) Time after Time eradication no no no no no no no no no no no yes yes yes 3/11 3/11 Yes/no: Yes/no: Current drinking no no no no no no no no no no no no no no 0/14 Yes/no: Yes/no: Current smoking - - - - cancer LB-LC UB-LC LB-PW LB-PW Ant-LC MB-LC Ant-GC MB-PW MB-PW MB-PW MB-PW Ant-AW Ant-AW Ant-AW Ant-AW As above LB-LC & Ant-LC & MB-AW & MB-AW Location of primary gastric Table 1. Characteristics of enrolled patients who suffered from gastric cancer at least 2 years after H. pylori eradication. 1. Characteristics of enrolled patients who suffered Table ).The macroscopic type of gastric cancer was classified according to the Japanese Classification of Gastric carcinoma. PG, pepsinogen; GC, gastric cancer; . d . s GC GC GC GC GC GC GC GC GC GC GU AG GU DU prior to GI Disease eradication DU:1, AG: 1 GC:10, GU:2, 64 66 60 72 74 80 76 80 76 78 81 74 70 71 48 (9) age F F F F M M M M M M M M M M Sex 10/4 M/F: Deta represent mean ( GU, gastric ulcer; AG, atrophic gastritis; DU, duodenal ulcer; UB, upper body; MB, middle body; LB, lower body; Ang, angularis; Ant, antrum; GC, greater curvature; LC, lesser curva - Ang, angularis; AG, atrophic gastritis; DU, duodenal ulcer; UB, upper body; MB, middle LB, lower GU, gastric ulcer; AS, acid-secreting area. posterior wall; M, mucosal infiltration; SM, submucosal NAS, non-acid-secreting area; anterior wall; PW, AW, ture; 2 1 3 4 5 6 7 8 9 11 10 13 12 14 No. Total Total Patient 48 K. Iijima et al. for the site-directed analysis. The diameters of the cancer- corpus including the angularis and the remaining 6 on the ous lesions ranged from 5 to 20 mm, averaging 10 mm, and antrum. No lesions involved the cardia (Table 1). The when dividing the neoplastic lesion into the intestinal- and greater curvature of the corpus escaped from the post-eradi- diffuse-type according to the Lauren classification (Laurén cation carcinogenesis, while there was no such tendency in 1965), all lesions in this study were defined as the intesti- the antrum. nal-type. The histological evaluation of the endoscopically Congo-red chromoendoscopy revealed the predomi- and surgically resected specimens revealed that all 19 nance of acid-secreting areas in the greater curvature over lesions were superficial cancers. The serum pepsinogen I those of the lesser curvature (Table 2). That is, while the concentration was less than 50 ng/ml in 9 of 11 available whole or the majority of the greater curvature was occupied subjects, indicating the residual extensive fundic atrophy in by acid-secreting areas in 10 of 14 subjects, there was no these patients. Regarding the locations of the 19 cancerous acid-secreting area observed in the lesser curvature in 8 of lesions within the stomach, 13 lesions were located on the 14 subjects. Additionally, there was no acid-secreting area anywhere in the whole stomach, representing complete achlorhydria, in 2 subjects, one of whom was found to be Table 2. Semi-quantitative assessment of the extent of anti-intrinsic factor-positive in the subsequent blood test acid-secreting areas in the lesser or greater curva- and judged to be complicated by type-A autoimmune gas- ture. tritis. Patient No. Lesser curvature Greater curvature Regarding the distribution of cancer development 1 None Intermediate within the stomach in relation to the regional acid-secreting capacity, all 19 cancerous lesions arose exclusively from 2 None Large non-acid-secreting areas (Table 1). Notably, the finding that 3 Small All 13 lesions were located on the corpus including the angula- 4 None Large ris implies that those lesions arose from irreversibly dys- 5 None Large functional mucosa after eradication, considering that the 6 None Intermediate acid secretory capacity is an intrinsic property of these ana- 7 None None tomical sites in subjects who have never been H. pylori- 8 Small All positive. Fig. 1 shows a representative patient of a patient 9 None Large (74 y.o., male) who suffered from double superficial gastric 10 Large All cancer detected in the posterior wall of the upper body 13 11 Large All years after eradication. The Congo-red chromoendoscopy 12 Small All of this patient revealed that both of the cancerous lesions 13 none none arose from non-acid-secreting areas. Fig. 2 shows another 14 large all representative patient (81 y.o., female) of a new, minute, Acid-secreting mucosa was assessed by Congo-red gastric cancer detected 14 years after eradication. She chromoendoscopy, and the area was semiquantitatively underwent Congo-red chromoendoscopy prior to eradica- classified into 4 groups in the lesser curvature and 5 groups tion as described in our previous study (Sekine et al. 2006; in the greater curvature, respectively. Iijima et al. 2009a). Although non-acid-secreting areas had

Fig. 1. Representative endoscopic images of patient 5. Shown are the representative endoscopic images of routine endoscopy (a) and Congo-red chromoendoscopy (b) of patient 5. a) 0-IIc type gastric cancer was detected in the posterior wall of the upper body (arrowhead). In the vicinity of the lesion, another minute cancer lesion was also identified (arrow). b) Congo-red chromoendoscopy revealed that both lesions arose from non color-shifting (non-acid-secreting) areas. Gastric Cancer Emerging after Eradication 49

Fig. 2. Representative endoscopic images of patient 1. Shown are the representative endoscopic images of patient 11 prior to (a), 6 months after (b), and 14 years after H. pylori eradication (c, d). a): Congo-red chromoendoscopy prior to eradication performed in 1996 revealed that the non-acid-secreting-area spread in the lesser curvature, anterior wall, and posterior wall of the fundus, while the acid-secreting existed only in the greater curvature. b) Congo-red chromoendoscopy performed 7 months after eradication showed that the acid-secreting area had expanded and the non-acid-secreting area remained only in the lower body. c) A minute gastric cancer (arrow) emerged 14 years after eradication in the anterior wall of the angularis. A single coagulation marker (arrowhead) was used to help distinguish the lesion. d) Congo-red chromoendoscopy at that time clearly revealed that the cancerous lesion arose from the non-acid-secreting-area. been predominant in her stomach prior to eradication (Fig. scores between the two areas was significant (P < 0.01). 2A), acid-secreting areas expanded so remarkably by 7 There was no neutrophil infiltration observed in any biopsy months after eradication that non-acid-secreting areas specimens (activity score = 0). remained in only a small portion of the lower body (Fig. Immunohistochemistry for Ki-67 as a marker for cell 2B). Thereafter, she was followed with endoscopic exami- proliferation revealed that there were remarkable (approxi- nation once a year, and a minute gastric cancer was eventu- mately 6 fold) differences in the protein expression between ally detected 14 years after eradication (Fig. 2C). Congo- the two areas (P < 0.01, Fig. 4C); marked Ki-67 positivity red chromoendoscopy at that time clearly revealed that the was diffusely observed in the epithelium of non-acid-secret- cancerous lesion arose from the residual non-acid-secreting- ing areas (Fig. 4A), while the protein expression was area (Fig. 2D). detected only in the nuclei of the cells of the glandular neck Histologically, there were striking differences in the in acid-secreting areas (Fig. 4C). Likewise, p53 immuno- grade of mucosal atrophy and intestinal metaplasia between histochemistry was also significantly different between the the acid-secreting and non-acid-secreting areas (P < 0.01 two areas (P < 0.01, Fig. 5C). Substantial nuclear expres- for both parameters). While the non-acid-secreting areas sion of p53 was identified in the epithelium of the non-acid- showed marked degrees of residual mucosal atrophy and secreting areas (Fig. 5A), while positivity was rarely intestinal metaplasia, the acid-secreting areas were free of observed in acid-secreting areas (Fig. 5B). such histological changes in the majority of cases (Fig. 3). In addition, a substantial degree of residual inflammation Discussion was also found in the non-acid-secreting areas in most Using Congo-red chromoendoscopy, we have already cases, whereas only mild inflammation was detected in shown that acid-secreting areas were significantly expanded acid-secreting areas, and the difference in inflammation shortly after H. pylori eradication, mainly with the resolu- 50 K. Iijima et al.

tion of inflammation (Sekine et al. 2006). However, the restoration was incomplete in that residual, non-acid-secreting areas remained in the fundus even in the long-term follow-up (Iijima et al. 2009a). The present study extended our previous observation by showing that gastric cancer newly emerged after the eradication exclusively from the residual, non-acid-secreting areas. Therefore, such non-acid-secreting areas identified by Congo-red chromoendoscopy could be regarded as the major carcinogenic background for gastric cancer arising after H. pylori eradication. There have been several reports that showed the incidence and characteristic features of gastric cancer after H. pylori eradication (Take et al. 2005; Kamada et al. 2005; Takenaka et al. 2007; Fukase et al. 2008; Ogura et al. 2008; Yanaoka et al. 2009; de Vries et al. 2009). However, many of these included cases whose cancer lesions developed relatively soon after eradication, namely within 2 years after Fig. 3. Differences in histological findings between acid- secreting and non-acid-secreting areas. the treatment (Take et al. 2005; Kamada et al. 2005; Biopsy samples from acid-secreting and non-acid-secret- Takenaka et al. 2007; Fukase et al. 2008; Ogura et al. 2008). ing areas of each subject were semi-quantitatively evalu- In the present study, we defined post-eradication gastric ated regarding inflammation, atrophy, and intestinal cancers as lesions that were discovered at least 2 years after metaplasia. In two subjects, there was no-acid-secreting area in their stomachs and the data only from the non- the eradication to minimize the possibility of mistaking a acid-secreting areas were plotted. NAS: non-acid-secret- minute cancer that had already existed at the time of the ing areas; AS: acid-secreting-areas; IM: intestinal meta- eradication therapy. Consequently, during the long-term plasia. periodic follow-up, with a mean duration of more than 6 *represents statistical difference with p < 0.01. years, we detected 14 cases in which the discovered cancer

Fig. 4. Comparison of immunohistochemical analysis on Ki-67 between acid-secreting and non-acid-secreting areas. a) Immunohistochemical image of Ki-67 from the non-acid-secreting area showed diffuse Ki-67 positivity in the epithelium. b) Immunohistochemical image of Ki-67 from the acid-secreting area of the same subject showed only scattered positivity in the nuclei of the cells of glandular neck. c) A comparison of the labeling index for Ki-67 between the two areas showed large differences between them (P < 0.01). Note that the data for acid-secreting areas were not available in 2 subjects, because acid-secreting areas did not appear in these patients. NAS: non-acid-secreting area; AS: acid- secreting area. Gastric Cancer Emerging after Eradication 51

Fig. 5. Comparison by immunohistochemical analysis of p53 between acid-secreting and non-acid-secreting areas. a) immunohistochemical image of p53 from the non-acid-secreting area showed substantial nuclear expression of p53 in the epithelium. Arrows represent immuno-positive cells for p53. b) immunohistochemical image of p53 from acid- secreting area of the same subject showed no expression of p53. c) A comparison of labeling index for p53 between the two areas showed significant differences between them (P < 0.01). The data for acid-secreting areas were not available in 2 subjects, because acid-secreting areas did not appear in these patients. NAS: non-acid-secreting area; AS: acid- secreting area. lesions were still tiny, with a mean diameter of 10 mm, pylori eradication showed extensive residual inflammation, strongly suggesting that the lesions of our present case mucosal atrophy, intestinal metaplasia, and sustained hyper- series developed de novo after eradication. proliferation compared with acid-secreting areas. In the Before the recognition of H. pylori infection in a present study of patients who suffered from gastric cancer human stomach, Tatsuta et al (Tatsuta et al. 1979) found, after eradication, we found similar differences in the histo- using Congo-red chromoendoscopy, that gastric cancer logical features between the two functionally distinct areas, developed chiefly from non-acid-secreting areas and, ina but the differences between the two areas were more promi- particular intestinal type of cancer, invariably from such nent in the patients of the present study compared with areas, and that such functionally impaired areas were char- those of the previous study (e.g. for Ki-67 immunostaining, acterized by mucosal inflammation as well as mucosal atro- 6 fold difference between the 2 areas in the former study vs. phy. The present study extends this previous observation 2.5 fold in the latter (Iijima et al. 2009a)). Such severe his- by showing that, although the non-acid-secreting areas are tological abnormalities persisting after eradication could considerably reduced by the elimination of H. pylori (Sekine predispose some patients to gastric carcinogenesis. et al. 2006; Iijima et al. 2009a), gastric carcinoma of the It has been reported that immunohistochemical accu- intestinal type still arises from residual, non-acid-secreting mulation of p53 was found in gastric mucosa infected with areas after eradication. We also found that such non-acid- H. pylori infection and that it significantly decreased after secreting areas were mainly characterized by marked muco- eradication (Jones et al. 1997; Nardone et al. 1999; Satoh et sal atrophy and intestinal metaplasia. Regarding the distri- al. 2001; Kodama et al. 2007). On the other hand, another bution of gastric cancer within the stomach, the greater study pointed out that significantly higher accumulation of curvature of the fundus and angularis was completely free p53 was detected in the gastric mucosa of patients with of cancer, which might reflect our previous observation that multifocal atrophic gastritis 1 year after eradication com- acid secretion is more easily restored after eradication in pared with that of patients with non-atrophic superficial the greater curvature compared to the lesser curvature gastritis (Guarner et al. 2005). Partly consistent with this (Sekine et al. 2006). study, the present study clearly revealed a site-specific dif- In our previous long-term follow-up study of non-can- ference in the p53 accumulation in H. pylori-eradicated cer patients (Iijima et al. 2009a), we already reported that patients, that is, sustained expression of p53 protein was the non-acid-secreting gastric mucosa remaining after H. observed in non-acid-secreting areas, while the protein was 52 K. Iijima et al. rarely expressed in the neighboring acid-secreting areas, fundic atrophy, the re-emergence of the normal fundic fine suggesting that accumulation of the protein could persist structure (fundic pit) after eradication could be detected by after eradication in non-acid-secreting areas. Previous stud- magnifying endoscopy (Yagi et al. 2005). This might sug- ies have indicated that over-expression of p53 protein gest that advanced endoscopy could make it possible to dis- detected by immunohistochemistry generally represents criminate between non-acid-secreting areas with high carci- mutant p53 (Nardone et al. 1999; Murakami et al. 1999; nogenic potential from the remaining acid-secreting areas. Kodama et al. 2007), suggesting that the immunochisto- chemical accumulation of p53 is an indicator of the loss of Acknowledgments the p53 tumor suppressor function. Therefore, the p53 pro- This work was supported in part by a Grant-in-Aid to K. I. tein accumulation detected in non-acid-secreting areas from JFE (The Japanese Foundation for Research and Promotion might represent important molecular events involved in the of Endoscopy) Grant. persistent gastric carcinogenesis after H. pylori eradication. The induction of fundic gastric cancer post-eradication Conflict of Interest exclusively from the functionally irreversible, non-acid- All authors have no conflict of interest in this study. secreting areas could suggest that such areas had already progressed to an irreversible stage of histological change References leading to genetic alterations of genes that are crucial to Bansal, A., Ulusarac, O., Mathur, S. & Sharma, P. (2008) Correla- gastric carcinogenesis, such as p53. Thus, the progress tion between narrow band imaging and nonneoplastic gastric from atrophy and intestinal metaplasia to cancer might be pathology: a pilot feasibility trial. Gastrointest. Endosc., 67, an autonomous, H. pylori-independent process (de Vries et 210-216. de Vries, A.C., Kuipers, E.J. & Rauws, E.A. (2009) Helicobacter al. 2009). Simultaneously, the slightly but significantly pylori eradication and gastric cancer: when is the horse out of greater residual inflammation in the non-acid-secreting the barn? Am. J. Gastroenterol., 104, 1342-1345. areas could also be responsible for the carcinogenesis in the El-Omar, EM., Oien, K., El-Nujumi, A., Gillen, D., Wirz, A., areas after eradication through the sustained pro-carcino- Dahill, S., Williams, C., Ardill, J.E. & McColl, K.E. (1997) Helicobacter pylori infection and chronic gastric acid hypose- genic actions of sustained inflammatory cytokines (Gonda cretion. Gastroenterology, 113, 15-24. et al. 2009). Fukase, K., Kato, M., Kikuchi, S., Inoue, K., Uemura, N., One advantage of the present study is the enrollment Okamoto, S., Terao, S., Amagai, K., Hayashi, S. & Asaka, M. of patients with superficial cancers so that we could pin- (2008) Effect of eradication of Helicobacter pylori on inci- point the origin of the cancer whether they arose from acid- dence of metachronous gastric carcinoma after endoscopic resection of early gastric cancer: an open-label, randomised secreting or non-acid-secreting mucosa. This is an impor- controlled trial. Lancet, 372, 392-397. tant point since cancer lesions often arose in the vicinity of Gonda, T.A., Tu, S. & Wang, T.C. (2009) Chronic inflammation, the boundary of the two areas. On the other hand, a poten- the tumor microenvironment and carcinogenesis. Cell Cycle, tial limitation is the small numbers of the subjects enrolled. 8, 2005-2013. Review. Guarner, J., Bartlett, J., Seitz, R., Whistler, T., Herrera-Goepfert, Since the occurrence of gastric cancer after H. pylori eradi- R., Mohar, A., Sanchez, L., Halperin, D. & Parsonnet, J. cation is still relatively scarce, subjects were strictly (2005) Cell proliferation and inflammation on biopsy samples screened by enrolling only those in whom gastric cancer with multifocal atrophic gastritis before and 1 year after Heli- was discovered at least 2 years after the eradication, which cobacter pylori eradication. Arch. Pathol. Lab. Med., 129, 1451-1456. prevented us from enrolling more patients despite the Iijima, K., Sekine, H., Koike, T., Imatani, A., Ohara, S. & Shimose- 5-year enrollment period. Another limitation is that all can- gawa, T. (2004) Long-term effect of Helicobacter pylori erad- cer lesions in the present analysis were the intestinal type. ication on the reversibility of acid secretion in profound hypo- Although the present results on the cancer histology are in chlorhydria. Aliment. Pharmacol. Ther., 19, 1181-1188. Iijima, K., Koike, T., Sekine, H., Abe, Y., Asanuma, K., Ara, N., agreement with a recent review describing that most gastric Uno, K., Imatani, A., Ohara, S. & Shimosegawa, T. (2009a) cancers that developed after eradication were the intestinal- Sustained epithelial proliferation in a functionally irreversible type (Ito et al. 2009), further studies on the diffuse-type of fundic mucosa after Helicobacter pylori eradication. J. gastric cancer after H. pylori eradication are required. Gastroenterol., 44, 47-55. Iijima, K., Abe, Y., Kikuchi, R., Koike, T., Ohara, S., Sipponen, P. In conclusion, gastric cancer emerged de novo after & Shimosegawa, T. (2009b) Serum biomarker tests are useful eradication exclusively from non-acid-secreting areas. in delineating between patients with gastric atrophy and Because this study employing Congo-red chromoendoscopy normal, healthy stomach. World J. Gastroenterol., 15, confirms the significance of non-acid-secreting areas as the 853-859. Ito, M., Takata, S., Tatsugami, M., Wada, Y., Imagawa, S., Matsu- origin of carcinogenesis after eradication, further trials for moto, Y., Takamura, A., Kitamura, S., Matsuo, T., Tanaka, S., the identification of such high risk areas with recently Haruma, K. & Chayama, K. (2009) Clinical prevention of advanced endoscopy, such as high-resolution endoscopy gastric cancer by Helicobacter pylori eradication therapy: a with narrow band imaging (NBI) (Uedo et al. 2006; Bansal systematic review. J. Gastroenterol., 44, 365-371. Review. et al. 2008), should be a promising approach for estimating Jones, N.L., Shannon, P.T., Cutz, E., Yeger, H. & Sherman, P.M. (1997) Increase in proliferation and apoptosis of gastric the individual cancer risk in subjects after eradication. epithelial cells early in the natural history of Helicobacter Interestingly, a recent study has shown that, in patients with pylori infection. Am. J. Pathol., 151, 1695-1703. Gastric Cancer Emerging after Eradication 53

Kamada, T., Hata, J., Sugiu, K., Kusunoki, H., Ito, M., Tanaka, S., Gastrointest. Endosc., 64, 678-685. Inoue, K., Kawamura, Y., Chayama, K. & Haruma, K. (2005) Take, S., Mizuno, M., Ishiki, K., Nagahara, Y., Yoshida, T., Yokota, Clinical features of gastric cancer discovered after successful K., Oguma, K., Okada, H. & Shiratori, Y. (2005) The effect of eradication of Helicobacer pylori: Results from a 9-year eradicating helicobacter pylori on the development of gastric prospective follow-up study in Japan. Aliment. Pharmacol. cancer in patients with peptic ulcer disease. Am. J. Gastroen- Ther., 21, 1121-1126. terol., 100, 1037-1042. Kodama, M., Murakami, K., Okimoto, T., Sato, R., Watanabe, K. Takenaka, R., Okada, H., Kato, J., Makidono, C., Hori, S., & Fujioka, T. (2007) Expression of mutant type-p53 products Kawahara, Y., Miyoshi, M., Yumoto, E., Imagawa, A., in H pylori-associated chronic gastritis. World J. Gastroen- Toyokawa, T., Sakaguchi, K. & Shiratori, Y. (2007) Helico- terol., 13, 1541-1546. bacter pylori eradication reduced the incidence of gastric Kohda, K., Kuga, T., Kogawa, K., Kanisawa, Y., Koike, K., cancer, especially of the intestinal type. Aliment. Pharmacol. Kuroiwa, G., Hirayama, Y., Sato, Y. & Niitsu, Y. (2002) Effect Ther., 25, 805-812. of Helicobacter pylori eradication on platelet recovery in Tatsuta, M., Saegusa, T. & Okuda, S. (1973) Studies on gastritis in Japanese patients with chronic idiopathic thrombocytopenic upper portion of stomach by endoscopic Congo-red test. purpura and secondary autoimmune thrombocytopenic Endoscopy, 5, 61-69. purpura. Br. J. Haematol., 118, 584-588. Tatsuta, M., Okuda, S., Taniguchi, H. & Tamura, H. (1979) Gross Laurén, P. (1965) The two histological main types of gastric carci- and histological types of early gastric carcinomas in relation noma: diffuse and so called intestinal type carcinoma. An to the acid-secreting area. Cancer, 43, 317-321. attempt at a histo clinical classification. Acta. Pathol. Micro- Tucci, A., Poli, L., Tosetti, C., Biasco, G., Grigioni, W., Varoli, O., biol. Scand., 64, 31-49. Mazzoni, C., Paparo, G.F., Stanghellini, V. & Caletti, G. Misiewicz, J.J., Tytgat, G.N. & Goodwin, C.S. (1990) The Sydney (1998) Reversal of fundic atrophy after eradication of Helico- System; a new classification of gastritis. Working Party bacter pylori. Am. J. Gastroenterol., 93, 1425-1431. Reports, 1-10. Uedo, N., Ishihara, R., Iishi, H., Yamamoto, S., Yamamoto, S., Murakami, K., Fujioka, T., Okimoto, T., Mitsuishi, Y., Oda, T., Yamada, T., Imanaka, K., Takeuchi, Y., Higashino, K., Nishizono, A. & Nasu, M. (1999) Analysis of p53 gene muta- Ishiguro, S. & Tatsuta, M. (2006) A new method of diag- tions in Helicobacter pylori-associated gastritis mucosa in nosing gastric intestinal metaplasia: narrow-band imaging endoscopic biopsy specimens. Scand. J. Gastroenterol., 34, with magnifying endoscopy. Endoscopy, 38, 819-824. 474-477. Uemura, N., Okamoto, S., Yamamoto, S., Matsumura, N., Nardone, G., Staibano, S., Rocco, A., Mezza, E., D'armiento, F.P., Yamaguchi, S., Yamakido, M., Taniyama, K., Sasaki, N. & Insabato, L., Coppola, A., Salvatore, G., Lucariello, A., Figura, Schlemper, R.J. (2001) Helicobacter pylori infection and the N., De Rosa, G. & Budillon, G. (1999) Effect of Helicobacter development of gastric cancer. N. Engl. J. Med., 345, pylori infection and its eradication on cell proliferation, DNA 784-789. status, and oncogene expression in patients with chronic Yabuki, N., Sasano, H., Tobita, M., Imatani, A., Hoshi, T., Kato, K., gastritis. Gut, 44, 789-799. Ohara, S., Asaki, S., Toyota, T. & Nagura, H. (1997) Analysis Ogura, K., Hirata, Y., Yanai, A., Shibata, W., Ohmae, T., Mitsuno, of cell damage and proliferation in Helicobacter pylori- Y., Maeda, S., Watabe, H., Yamaji, Y., Okamoto, M., Yoshida, infected human gastric mucosa from patients with gastric H., Kawabe, T. & Omata, M. (2008) The effect of Helico- adenocarcinoma. Am. J. Pahtol., 151, 321-329. bacter pylori eradication on reducing the incidence of gastric Yagi, K., Honda, H., Yang, J.M. & Nakagawa, S. (2005) Magni- cancer. J. Clin. Gastroenterol., 42, 279-283. fying endoscopy in gastritis of the corpus. Endoscopy, 37, Satoh, K., Kihira, K., Kawata, H., Tokumaru, K., Kumakura, Y., 660-666. Ishino, Y., Kawakami, S., Inoue, K., Kojima, T., Satoh, Y., Yanaoka, K., Oka, M., Ohata, H., Yoshimura, N., Deguchi, H., Mutoh, H. & Sugano, K. (2001) p53 expression in the gastric Mukoubayashi, C., Enomoto, S., Inoue, I., Iguchi, M., mucosa before and after eradication of Helicobacter pylori. Maekita, T., Ueda, K., Utsunomiya, H., Tamai, H., Fujishiro, Helicobacter, 6, 31-36. M., Iwane, M., Takeshita, T., Mohara, O. & Ichinose, M. Sekine, H., Iijima, K., Koike, T., Abe, Y., Imatani, A., Kato, K., (2009) Eradication of Helicobacter pylori prevents cancer Ohara, S. & Shimosegawa, T. (2006) Regional differences in development in subjects with mild gastric atrophy identified the recovery of gastric acid secretion after Helicobacter pylori by serum pepsinogen levels. Int. J. Cancer, 125, 2697-2703. eradication: evaluations with Congo-red chromoendoscopy.