neutrophil infiltration and the distribution of intestinal metaplasia is associated with metachronous gastric cancer following endoscopic submucosal dissection

1Department of Gastroenterology, Kyoundo Hospital; 2Department of Gastroenterology, Graduate School of Medicine, the University of Tokyo; 3Department of Gastroenterology, Tokyo Metropolitan Police Hospital; 4Department of Gastroenterology, JR Tokyo General Hospital, Tokyo, Japan Correspondence and reprints: Dr Takafumi Sugimoto, Department of Gastroenterology, Graduate School of Medicine, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8655, Japan. Telephone 81-3-3815-5411 ext 33070, fax 81-3-3814-0021, e-mail sugimot-tky@umin.ac.jp Received for publication March 4, 2015. Accepted March 7, 2015 The rate of diagnosis of early gastric cancer has increased due to improved diagnostic procedures such as endoscopic examinations including chromoendoscopy and narrow-band imaging (1,2). Endoscopic resection (ER), which includes endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD), has been established as a less-invasive treatment for early gastric cancer without concomitant lymph node metastasis (3-5). ESD has become accepted for en bloc resection, which is beneficial for curative treatment to avoid local recurrence (5-10). Although ER contributes to preservation of the majority of the stomach, metachronous cancers developing at other sites in the stomach may occur more frequently after ER than those after surgical partial gastrectomy (11). Several studies have reported that the incidence of metachronous cancers after ER is 2.4% to 14% (12,13). The residual gastric mucosa after ER is believed to be a high-risk microenvironment. Helicobacter pylori is one of the most important risk factors for gastric cancer (14-17). Inflammation caused by H pylori infection may play an initiating role in gastric carcinogenesis (18). Based on this hypothesis, intestinal-type gastric adenocarcinoma develops through multistep mucosal changes from superficial gastritis to atrophic T Sugimoto, Y Yamaji, K Sakitani, et al. Neutrophil infiltration and the distribution of intestinal metaplasia is associated with metachronous gastric cancer following endoscopic submucosal dissection. Can J Gastroenterol Hepatol 2015;29(6):321-325.

T he rate of diagnosis of early gastric cancer has increased due to improved diagnostic procedures such as endoscopic examinations including chromoendoscopy and narrow-band imaging (1,2). Endoscopic resection (ER), which includes endoscopic mucosal resection (EMR) and endoscopic submucosal dissection (ESD), has been established as a less-invasive treatment for early gastric cancer without concomitant lymph node metastasis (3)(4)(5). ESD has become accepted for en bloc resection, which is beneficial for curative treatment to avoid local recurrence (5)(6)(7)(8)(9)(10). Although ER contributes to preservation of the majority of the stomach, metachronous cancers developing at other sites in the stomach may occur more frequently after ER than those after surgical partial gastrectomy (11). Several studies have reported that the incidence of metachronous cancers after ER is 2.4% to 14% (12,13). The residual gastric mucosa after ER is believed to be a high-risk microenvironment.
Helicobacter pylori is one of the most important risk factors for gastric cancer (14)(15)(16)(17). Inflammation caused by H pylori infection may play an initiating role in gastric carcinogenesis (18). Based on this hypothesis, intestinal-type gastric adenocarcinoma develops through multistep mucosal changes from superficial gastritis to atrophic   gastritis, intestinal metaplasia (IM) and dysplasia (18). Eradication of H pylori is a highly useful method of preventing primary and metachronous gastric cancers (19)(20)(21). We have reported previously that IM and neutrophil infiltration (NI) are high-risk microenvironments and risk factors for primary gastric cancers (22). However, the risk factors for metachronous cancers are unclear. Thus, the aim of the present study was to evaluate the risk factors for metachronous gastric cancers after ESD in a long-term follow-up.

METHODS Patients
A total of 201 consecutive patients were treated using ESD at the University of Tokyo Hospital (Tokyo, Japan) between September 2000 and September 2009. All demographic data and endoscopic findings obtained before ESD were evaluated retrospectively from patient records. The location, macroscopic types and histological findings of the tumours were categorized according to the Japanese classification of gastric carcinoma (23). Expanded ESD criteria have been proposed by Gotoda et al (24). All patients were followed up for at least one year. Forty-six patients were excluded from the study: 19 did not meet the ESD specimen histology criteria and required additional surgery, five had remnant stomach cancers, nine were followed for up to one year, four did not have H pylori information and, in nine patients, ESD was performed before enrollment. The present study was approved by the Institutional Review Board of the University of Tokyo.

Disease assessment
Experienced endoscopists performed the gastrointestinal endoscopy procedures, and gastric cancer was diagnosed by histology. Biopsy specimens were obtained from the greater curvature of the antrum and middle corpus to evaluate gastric mucosal status including H pylori, IM and NI before ESD (22). Both a rapid urease test (Helicocheck, Otsuka Pharmaceuticals, Japan) and histopathological examination were performed. A positive result on at least one test was deemed to be evidence of H pylori infection. Blood samples were also collected before ESD to measure serum levels of pepsinogen I and II using ELISA kits (LS test Eiken Kagaku, Inc, Japan). ESD was performed as described previously (8,9). Follow-up endoscopy after ESD for detection of metachronous gastric cancers was scheduled at two and six months, one year and annually thereafter. Lesions detected within one year of initial ESD were regarded as synchronous multiple lesions because microcancers may have been missed at the time of ESD (12,25). Metachronous gastric cancers were defined as new gastric cancers in different areas from the initial lesion and occurring at least one year after the initial ESD. Eradication of H pylori was recommended when possible. Metachronous gastric cancer was defined as the development of a new carcinoma in areas other than the primary gastric cancer site at least one year after ESD. H pylori-positive patients were treated with triple therapy consisting of 200 mg clarithromycin, 750 mg amoxicillin and 30 mg lansoprazole twice daily for one week after ER. Patients in whom H pylori was not eradicated were treated with second-line therapy consisting of 250 mg metronidazole instead of clarithromycin (26). Eradication was confirmed by a negative 13 C-urea breath test (8,9).

Statistical analyses
Study subjects were categorized according to the presence of the metachronous gastric cancer. All analyses were performed using JMP version 9 (SAS Institute, USA). Student's t tests were used for intergroup comparison of mean age, follow-up period and tumour size. The other patient clinical characteristics and histopathological characteristics of the gastric mucosa and cancers were compared using the χ 2 test or Fisher's exact test, as appropriate. Relative risks for metachronous gastric cancer were calculated using the Cox proportional hazards model. Differences were considered to be statistically significant at the 5% probability level.

Baseline characteristics of the study subjects
Baseline clinical characteristics of the patients are summarized in Table 1. A total of 155 patients (119 men, 36 women, mean age 68.9 years) were followed for up to 10.8 years (mean 4.2 years). Six patients died during the study period, but none of the deaths were related to gastric cancer. Twenty-three (14%) patients developed metachronous gastric cancers (3.5% per year).

Factors related to metachronous gastric cancers
The cumulative incidence of metachronous gastric cancer is summarized in Table 2, and Figures 1 and 2. The ORs for metachronous gastric cancers in IM and NI in the corpus were 3.00 (P=0.010) and 3.47 (P=0.0034). IM and NI in the corpus and tumour size were risk factors. Successful H pylori eradication did not contribute significantly to the reduction of risk for

Multivariate analysis for metachronous gastric cancer
The multivariate analysis of tumour size, IM and NI in the corpus showed that IM and NI in the corpus were independent risk factors for metachronous gastric cancer ( Table 3). The ORs for metachronous gastric cancer were 2.65 (95% CI 1.13 to 6.66; P=0.024) in the IM-positive group in the corpus and 3.06 (95% CI 1.32 to 7.64; P=0.0091) in the NI-positive group in the corpus.

DISCUSSION
The results of the present single-centre study involving 155 consecutive patients showed that IM and NI in the corpus were related to metachronous gastric cancer. IM and NI were high-risk microenvironments and risk factors for primary and metachronous gastric cancers. Not only H pylori eradication but also annual follow-up endoscopy may be important for improving the prognosis of patients with metachronous gastric cancer. Although many studies of residual gastric cancer have been conducted (27,28), few reports of metachronous gastric cancers after ESD for early gastric cancer are available. The rate of recurrence of early gastric cancer in the gastric stump has been estimated to be 1% to 3% (13,29). An increased recurrence rate would be expected in patients treated with ESD, in proportion to the larger area of gastric mucosa remaining in these patients. Other studies have reported that the annual incidence of metachronous gastric cancer after EMR is 2.5% to 4% (12,20). In the present study, the annual incidence was approximately 3.5%.
In their randomized controlled study, Fukase et al (20) reported that eradicating H pylori reduced the risk for developing metachronous gastric cancer in patients treated with EMR. In our previous study, we reported that IM, NI and the gastritis pattern are related to developing primary gastric cancer (22). In the present study, our data revealed that IM and NI were also related to metachronous gastric cancer; however, no significant difference in H pylori status was observed. It has been suggested that a larger sample size would demonstrate a significant reduction in the incidence of metachronous gastric cancer after eradicating H pylori. Some studies have reported a significant reduction in cancer development following H pylori eradication in subgroups with mild atrophy or antrum-dominant atrophy (30,31). The preventive effect of H pylori eradication against gastric cancer may depend on the degree of baseline gastric mucosal atrophy and IM. Many patients in our study had IM, which may represent a point of no return at which the development of gastric cancer can no longer be prevented by eradicating H pylori.
In a previous study, we reported that NI in the corpus was related to prevalent diffuse-type gastric cancer (22). In that study, the diffuse type of the metachronous cancers were found in only two cases. A prospective study with a larger number is needed to elucidate the difference between diffuse type and intestinal type cancers.
The annual incidence of metachronous gastric cancer in the present study was 3.5%. The cumulative prevalence of gastric remnant cancer after surgical partial gastrectomy for early gastric cancer is reported to be 2.4% at five years. Patients with early gastric cancer after ESD treatment have been identified as a high-risk group for metachronous gastric cancer. If H pylori is eradicated, the NI status of gastric mucosa improves (32). However, IM of the gastric mucosa cannot easily be improved (33)(34)(35)(36). Endoscopists should recognize the characteristics of metachronous gastric cancer and special attention should be devoted to detection of early gastric cancer.
Six patients died during the study period, but none of these deaths were related to gastric cancer. All patients with metachronous gastric cancer were treated completely with ESD or surgery. Annual endoscopy follow-up may be useful for detection of metachronous gastric cancer during the early stage and prevention of gastric cancer death.
A limitation of our study may be a retrospective cohort study in consecutive patients. A prospective study is needed to determine whether baseline histological characteristics can be predictive factors for metachronous gastric cancer and be used to determine the intensity of surveillance endoscopy among this patient population.

CONCLUSION
The presence of IM and NI in the corpus was closely related to the development of metachronous gastric cancer after ESD.

DISCLOSURES:
The authors have no financial disclosures or conflicts of interest to declare.

ACKNOWLEDGEMENTS:
The authors thank the gastroenterologists at the University of Tokyo Hospital for their endoscopic procedure. They are grateful to all the pathologists for their histopathological examinations on our subjects.