Introduction. Chihiro Takeuchi 1 Nobutake Yamamichi. Yu Takahashi 1 Toru Mitsushima

Gastric Cancer (2017) 20:314 321 DOI 10.1007/s10120-016-0607-3 ORIGINAL ARTICLE Gastric polyps diagnosed by double-contrast upper gastrointestinal barium X-ray radiography mostly arise from the Helicobacter pylori-negative stomach with low risk of gastric cancer in Japan Chihiro Takeuchi 1 Nobutake Yamamichi 1 Takeshi Shimamoto 1,2 Yu Takahashi 1 Toru Mitsushima 2 Kazuhiko Koike 1 Received: 26 January 2016 / Accepted: 3 March 2016 / Published online: 14 March 2016 The International Gastric Cancer Association and The Japanese Gastric Cancer Association 2016 Abstract Background Double-contrast upper gastrointestinal barium X-ray radiography (UGI-XR) is a method broadly used for gastric cancer screening in Japan. Gastric polyp is one of the most frequent findings detected by UGI-XR, but how to handle it remains controversial. Methods Gastric polyps of the 17,264 generally healthy subjects in Japan who underwent UGI-XR or upper gastrointestinal endoscopy (UGI-ES) in 2010 were analyzed. Results Of the 6,433 UGI-XR examinees (3,405 men and 3,028 women, 47.4 ± 9.0 years old), gastric polyps were detected in 464 men (13.6 %) and 733 women (24.2 %) and were predominantly developed on the non-atrophic gastric mucosa (p \ 0.0001). Multiple logistic regression analysis showed that the presence of gastric polyps has C. Takeuchi and N. Yamamichi equally contributed to this work. & Nobutake Yamamichi nyamamic-tky@umin.ac.jp Chihiro Takeuchi goodjob951@gmail.com Takeshi Shimamoto tsimamoto@gmail.com Yu Takahashi yutakahashi-tky@umin.ac.jp Toru Mitsushima mitsushima@kameda.jp Kazuhiko Koike kkoike-tky@umin.ac.jp 1 2 Department of Gastroenterology, Graduate School of Medicine, The University of Tokyo, 7-3-1, Hongo, Bunkyoku, Tokyo 113-8655, Japan Kameda Medical Center Makuhari, CD-2, 1-3, Nakase, Mihama-ku, Chiba 261-0023, Japan significant association with lower value of serum anti- Helicobacter pylori IgG titer, female gender, lighter smoking habit, older age, and normal range of body mass index ( 18.5 and \25), but not with drinking or serum pepsinogen I/II ratio. During the 3-year follow-up, gastric cancer occurred in 7 subjects (0.11 %), but none of them had gastric polyps at the beginning of the follow-up period. Of the 2,722 subjects with gastric polyps among the 10,831 UGI-ES examinees in the same period, 2,446 (89.9 %) had fundic, 267 (9.8 %) had hyperplastic, and 9 (0.3 %) had adenomatous/cancerous polyps. Conclusions Gastric polyps diagnosed by UGI-XR predominantly arise on the Helicobacter pylori-negative gastric mucosa with a low risk of gastric cancer in Japan. In the prospective observation, none of the UGI-XR examinees with gastric polyps developed gastric cancer for at least 3 years subsequently. Keywords Gastric polyp Double-contrast upper gastrointestinal barium X-ray radiography (UGI-XR) Upper gastrointestinal endoscopy (UGI-ES) Helicobacter pylori Gastric cancer Introduction Double-contrast upper gastrointestinal barium X-ray radiography (UGI-XR) is the most widely used method for gastric cancer screening in Japan [1]. Since the 1960s, the UGI-XR-based gastric cancer screening program endorsed by the Japanese government has achieved a significant reduction in mortality and morbidity of gastric cancer [1 3]. UGI-XR can detect not only gastric cancer but also gastroduodenal erosion or ulcer, gastric or esophageal polyps, esophagogastric diverticula, atrophic or

Gastric polyps diagnosed by double-contrast upper gastrointestinal barium X-ray radiography 315 hypertrophic gastritis, esophageal hiatal hernia, and other lesions [4, 5]. Of these various lesions, gastric polyp is one of the most frequent findings observed in the usual screening examination [6]. Nowadays, however, it is a matter of controversy how to manage the gastric polyps detected by UGI-XR. Actually, the clinical steps after detecting gastric polyps by UGI-XR are quite varied in Japan. Some UGI-XR examinees diagnosed with gastric polyps are recommended to undergo a detailed examination with upper gastrointestinal endoscopy (UGI-ES), whereas other examinees are advised to obtain a follow-up survey because the risk of developing gastric cancer is low. Gastric polyp is a mere morphological entity including protruded gastric mucosa of various histology [7 10], and the distribution of its histological types shows marked regional differences around the world [6, 11 13]. Recently, two international guidelines of gastric polyps were released from the United States (US) and UK [7, 9]. The former is a pathology-based guideline targeting both epithelial and nonepithelial lesions [9], which is mostly based on a large-scale nationwide survey in the US [8]. In contrast, the latter mostly focuses on epithelial lesions using the GRADE (grading of recommendations, assessment, development, and evaluation) system to develop a guideline consensus [7]. Regardless of the several differences, both guidelines recommend forceps biopsy for histological evaluation of gastric polyps. Performing biopsy is not difficult under UGI-ES, which can histologically evaluate such types of gastric polyps as fundic gland polyps, hyperplastic polyps, adenomatous/cancerous polyps, inflammatory fibroid polyps, and hamartomatous polyps [7 10]. Conversely, however, it is impossible to execute a biopsy in the procedure of UGI-XR: the aforementioned guidelines are not easily applied for UGI-XR-detected gastric polyps. Based on this background, the risk of future gastric tumorigenesis should be evaluated to determine what to do when gastric polyps are detected by UGI-XR. Several risk factors, such as chronic H. pylori infection [14, 15], male gender [16, 17], salt intake [18, 19], and smoking [20 23], adversely affect gastric tumorigenesis. It is now established that H. pylori infection is by far the strongest risk factor among them [14], and it is also established that H. pyloriinduced mucosal atrophy (atrophic gastritis) and enlarged gastric folds (hypertrophic gastritis) are both obvious predictive indicators for future gastric cancer incidence [24, 25]. During this half century, the prevalence of H. pylori infection has considerably decreased not only in Japan but also worldwide [26 29]. Accordingly, the prevalence of various H. pylori-related gastric polyps must have changed, but recent epidemiological data concerning gastric polyps are insufficient. In the present study, using the large-scale data of generally healthy subjects in Japan, we analyzed the characteristics of UGI-XR/UGI-ES-detected gastric polyps. We believe the present results can reflect the background factors and histological breakdown of gastric polyps in East Asia today. We are also convinced that our data will be useful for the strategy against gastric polyps not only those detected by UGI-XR but also those found by other medical examinations. Methods Study subjects The study participants were 20,773 asymptomatic generally healthy subjects who underwent a medical checkup in 2010 at our institute (Chiba-shi, Chiba, Japan). After excluding 892 subjects with insufficient data, 215 subjects with a history of gastrectomy, 1,517 subjects with past eradication of H. pylori, and 885 subjects using gastric acid suppressants (proton pump inhibitors or histamine H 2 -receptor antagonists), 17,264 eligible subjects were analyzed (Fig. 1). The total numbers of study subjects comprised 6,433 UGI-XR examinees and 10,831 UGI-ES examinees. Follow-up strategy All the study subjects were recommended to undergo an annual health checkup including gastric cancer screening. Data of UGI-XR and UGI-ES from the study participants were thoroughly obtained from 2010 to 2013. Although gastric cancer could be detected by UGI-XR or UGI-ES, the final diagnosis was confirmed by histopathology using the biopsy specimens. Diagnosis of gastric polyps and atrophic gastritis by upper gastrointestinal X-ray (UGI-XR) examination A method for UGI-XR was described in detail in our previous report [5]. Gastric polyps were diagnosed Fig. 1 Study recruitment flowchart

316 C. Takeuchi et al. independently by two medical doctors specialized in gastroenterology or radiology and one radiologic technologist specialized in double-contrast X-ray imaging. Synthetic diagnoses from the judgments of these three specialists were used for the analyses. UGI-XR-based atrophic gastritis was classified into four types according to our recent reports as follows [5, 25, 30]: no atrophic gastric mucosa (type A), mild atrophic gastritis (type B), moderate atrophic gastritis (type C), and severe atrophic gastritis (type D). Statistical methods Univariate analysis was performed with the presence of gastric polyp as a response variable and seven background factors (age, gender, serum anti-h. pylori antibody, ratio of serum pepsinogen I/II, BMI, smoking, and drinking) as explanatory variables. Simple logistic regression was used for univariate analyses in which p \ 0.05 was considered as statistically significant. Next, all the values were standardized, and the multiple logistic regression model was applied to calculate the standardized coefficients and odds ratios for the seven factors. A two-sided p value of less than 0.05 was considered statistically significant. To evaluate the association of the UGI-XR-detected gastric polyp with the four grade types of UGI-XR-based atrophic gastritis, the Cochran Armitage test for trend was applied. A Kaplan Meier curve was generated to assess the incidence of gastric cancer using JMP 10 software (SAS Institute), in which p \ 0.05 was considered as statistically significant according to the log-rank test. All statistical analyses were performed using SAS version 8.2 (SAS Institute, Cary, NC, USA). Results Positively associated factors for the UGI-XRdetected gastric polyps based on univariate and multivariate analyses The characteristics of the 6433 UGI-XR examinees focusing on the presence of gastric polyps are shown in Table 1. The univariate analyses revealed that age, gender, serum anti-h. pylori IgG, pepsinogen I/II ratio, BMI, and smoking status have significant association with the presence of gastric polyps. Our results indicated that the subjects diagnosed with gastric polyps by UGI-XR tend to be younger female subjects free from H. pylori infection and smoking habit, within the normal range of BMI, and having higher pepsinogen I/II ratio. The following multivariate analysis showed that positively associated factors for the presence of UGI-XR-based gastric polyps in order of significance are lower value of serum anti-h. pylori IgG, female gender, lighter smoking habit, older age, and normal range of BMI (Table 2). Judging from the values of standardized coefficients, noninfection with H. pylori denoted the strongest positive association with gastric polyps by far. Female gender and lighter smoking habit showed the second and third strongest associations with polyps. UGI-XR-detected gastric polyps currently predominantly develop on the gastric mucosa with no atrophic change in Japan In our recent reports, we demonstrated that UGI-XR can detect atrophic gastritis accurately [5, 30]: UGI-XR-based atrophy shows a strong and significant association with UGI-ES-based atrophy according to the Kimura Takemoto classification [31]. In the present study, we analyzed the association between the four grade categories of UGI-XRbased atrophic gastritis [5, 30] and the presence of gastric polyps (Table 3). In Japan today, UGI-XR-detected gastric polyps predominantly develop on the gastric mucosa with no atrophic change (p \ 0.0001), which is consistent with Table 2 revealing the significant positive association between the UGI-XR-detected gastric polyps and no infection with H. pylori. The presence of gastric polyps suggests the present low risk of future gastric cancer development in Japan At the end of the 3-year follow-up, 4919 subjects underwent UGI-XR or UGI-ES once or more often for gastric cancer screening. During this period, 7 subjects developed gastric cancer, but none of them had gastric polyps at the beginning of this study (Fig. 2a). Namely, our data indicate that all the gastric polyps detected by UGI-XR have not developed into gastric cancer for at least 3 years after the screening. The Kaplan Meier curves evaluating the future incidence of gastric cancer could not detect a significant difference between the subjects with or without gastric polyps (Fig. 2b), but our data plainly suggest that future canceration of UGI-XR-based gastric polyps rarely occurs in Japan at present. Characteristics of the gastric polyps detected by upper gastrointestinal endoscopy (UGI-ES) Because it is impossible to differentially diagnose the type of gastric polyps by UGI-XR, we analyzed the study participants who underwent UGI-ES in the same period (Fig. 1). Of the 10,831 UGI-ES examinees, 2,446 (22.6 %) had fundic (gland) polyps, 267 (2.5 %) had hyperplastic

Gastric polyps diagnosed by double-contrast upper gastrointestinal barium X-ray radiography 317 Table 1 Characteristics of the 6433 subjects, focusing on the presence of gastric polyps diagnosed by upper gastrointestinal barium X-ray radiography (UGI-XR) Factors Subjects with gastric polyps (n = 1197) Subjects without gastric polyps (n = 5236) Odds ratio (95 % CI) p value Age 46.6 ± 8.1 (27 79) 47.5 ± 9.0 (20 83) \0.001* \40 235 (17.7 %) 1092 (82.3 %) Reference C40 and \50 555 (21.5 %) 2024 (78.5 %) 1.27 (1.08 1.51) C50 and \60 323 (16.9 %) 1586 (83.1 %) 0.95 (0.79 1.14) C60 and \70 76 (13.6 %) 482 (86.4 %) 0.73 (0.55 0.97) C70 8 (13.3 %) 52 (86.7 %) 0.71 (0.34 1.52) Gender \0.001* Female 733 (24.2 %) 2295 (75.8 %) Reference Male 464 (13.6 %) 2941 (86.4 %) 0.49 (0.43 0.56) Anti-H. pylori IgG \0.001* \3 1167 (25.7 %) 3374 (74.3 %) Reference C3 and \10 11 (5.0 %) 207 (95.0 %) 0.15 (0.08 0.28) C10 19 (1.1 %) 1655 (98.9 %) 0.03 (0.02 0.05) Pepsinogen I/II ratio \0.001* [3 1181 (20.9 %) 4483 (79.1 %) Reference B3 and [2 7 (1.4 %) 508 (98.6 %) 0.05 (0.02 0.11) B 2 9 (3.5 %) 245 (96.5 %) 0.14 (0.07 0.27) BMI 22.4 ± 3.1 22.9 ± 3.4 \0.001* \18.5 86 (19.1 %) 365 (80.9 %) 0.95 (0.74 1.22) C18.5 and \25.0 895 (19.9 %) 3613 (80.1 %) Reference C25.0 216 (14.7 %) 1258 (85.3 %) 0.69 (0.59 0.81) Alcohol 0.097 Rarely drinking 503 (19.6 %) 2064 (80.4 %) Reference Usually drinking 694 (18.0 %) 3172 (82.0 %) 0.90 (0.79 1.02) Smoking \0.001* Lifelong nonsmoker 810 (23.1 %) 2701 (76.9 %) Reference Past habitual smoker 245 (15.1 %) 1377 (84.9 %) 0.59 (0.51 0.69) Current smoker 142 (10.9 %) 1158 (89.1 %) 0.41 (0.34 0.50) Single logistic regression was used for statistical evaluation, and the association of each background factor between subjects having gastric polyps and those not having gastric polyps was calculated respectively CI confidence interval, H. pylori Helicobacter pylori, BMI body mass index * The level of significance in all the univariate analyses was set at p value \0.05 polyps, 9 (0.08 %) had adenomatous/cancerous polyps, and 8,112 (74.9 %) had no gastric polyps (Table 4). The rates of fundic, hyperplastic, and adenomatous/cancerous polyps among the 2,722 subjects with gastric polyps were 89.9 % (2,446), 9.8 % (267), and 0.3 % (9), respectively: these indicate that most of the gastric polyps detected nowadays in Japan are fundic gland polyps with no risk of future tumorigenesis. Distribution of the types of gastric polyps in the present study is in accordance with the recent reports not only from East Asia [6] but also from the United States [8]. Detailed characteristics of the subjects focusing on the UGI-ES-based gastric polyps are shown in Table 4. Fundic polyps with no risk of future canceration mostly developed on the H. pylori-negative gastric mucosa (96.8 %; 2368 of 2446). Conversely, hyperplastic polyps with some risk of future canceration tended to develop in the subjects with H. pylori infection (80.1 %; 214 of 267). Adenomatous/cancerous polyps with much higher risk of future canceration also predominantly developed in the H. pyloripositive subjects (88.9 %; 8 of 9). Of the 10,831 UGI-ES examinees, gastric cancer lesions were detected in 12 subjects (0.11 %), which included 3 erosive, 2 ulcerative, 1 flat and protruded, and 6 depressed tumors (Table 4). The flat and protruded lesion (0 IIb and 0 IIa macroscopic type of early gastric cancer) may resemble benign gastric polyps to some extent, but the other 11 malignant lesions could not be mistaken for gastric polyps. Altogether, our results indicate that it is quite rare

318 C. Takeuchi et al. Table 2 Association between the presence of UGI-XRdetected gastric polyps and the seven factors based on the multivariate analysis Variables Standardized coefficients Odds ratio (95 % CI) p value Anti-H. pylori IgG (reference: <3) C10-1.565 0.21 (0.16 0.26) \0.0001* C3 and \10-0.358 0.70 (0.62 0.78) \0.0001* Male gender (reference: female) -0.281 0.76 (0.70 0.82) \0.0001* Smoking (reference: lifelong nonsmoker) Current smoker -0.269 0.76 (0.70 0.83) \0.0001* Past habitual smoker -0.118 0.89 (0.82 0.96) 0.0030* Age 0.135 1.14 (1.07 1.23) 0.0002* BMI (reference: 18.5 and <25.0) C25.0-0.083 0.92 (0.86 0.99) 0.0267* \18.5-0.071 0.93 (0.87 1.00) 0.0397* Alcohol (reference: rarely drinking) Usually drinking 0.034 1.03 (0.97 1.11) 0.3350 Pepsinogen I/II ratio (reference: >3) B3 and [2-0.037 0.96 (0.74 1.21) 0.762 B2 0.088 1.09 (0.92 1.27) 0.274 We evaluated age as continuous variables. Multiple logistic regression analysis was applied to calculate standardized coefficients and odds ratios for the seven variables CI confidence interval, H. pylori Helicobacter pylori, BMI body mass index * The level of significance was set at p \0.05; the seven variables are shown in order of the absolute values of standardized coefficients Table 3 Association between the presence of gastric polyps and the four grade categories of atrophic gastritis diagnosed by UGI-XR (upper gastrointestinal barium X-ray radiography) The four grades categories of UGI-XRbased atrophic gastritis UGI-XR-detected gastric polyp (?) UGI-XR-detected gastric polyp (-) Total No-atrophic gastric mucosa 1152 (25.6 %) 3345 (74.4 %) 4497 (100 %) Mild atrophic gastritis 15 (6.4 %) 219 (93.6 %) 234 (100 %) Moderate atrophic gastritis 12 (1.5 %) 810 (98.5 %) 822 (100 %) Severe atrophic gastritis 18 (2.0 %) 862 (98.0 %) 880 (100 %) Total 1197 (18.6 %) 5236 (81.4 %) 6433 (100 %) p value p \ 0.0001* * Cochran Armitage test for trend was applied for evaluation, in which p value \0.05 was statistically significant Fig. 2 a Association between the presence of upper gastrointestinal barium X-ray radiography (UGI-XR)-detected gastric polyps and the development of gastric cancer in the 3-year follow-up of 6433 study subjects. b Kaplan Meier curves to assess the 3-year incidences of gastric cancer focusing on the presence of gastric polyps, in which a p value less than 0.05 was considered as statistically significant according to the log-rank test

Gastric polyps diagnosed by double-contrast upper gastrointestinal barium X-ray radiography 319 Table 4 Characteristics of the 10,831 UGI-ES (upper gastrointestinal endoscopy) examinees classified into four groups based on the presence and types of gastric polyps Factors Subjects with fundic polyps (n = 2446) Subjects with hyperplastic polyps (n = 267) Subjects with adenomatous or cancerous polyps (n = 9) Subjects without gastric polyps (n = 8112) Age 49.9 ± 8.6 (23 84) 55.9 ± 7.9 (21 77) 56.1 ± 6.5 (48 69) 51.1 ± 9.6 (20 87) \40 279 (21.6 %) 10 (0.8 %) 0 (0.0 %) 1002 (77.6 %) C40 and \50 933 (27.3 %) 36 (1.1 %) 1 (0.03 %) 2444 (71.6 %) C50 and \60 899 (21.5 %) 139 (3.3 %) 6 (0.14 %) 3139 (75.1 %) C60 and \70 298 (17.7 %) 72 (4.3 %) 2 (0.12 %) 1311 (77.9 %) C70 37 (14.1 %) 10 (3.8 %) 0 (0.0 %) 216 (82.1 %) Gender Female 1372 (30.5 %) 92 (2.0 %) 1 (0.02 %) 3040 (67.5 %) Male 1074 (17.0 %) 175 (2.8 %) 8 (0.13 %) 5072 (80.2 %) Anti-H. pylori IgG 2.2 ± 8.8 (0.1 148.3) 56.5 ± 56.4 (0.2 300) 17.1 ± 14.6 (0.3 40.5) 21.5 ± 36.5 (0.1 300) \3 2297 (33.8 %) 39 (0.6 %) 1 (0.01 %) 4461 (65.6 %) C3 and \10 71 (12.6 %) 14 (2.5 %) 3 (0.53 %) 478 (84.6 %) C10 78 (2.2 %) 214 (6.2 %) 5 (0.14 %) 3173 (91.5 %) Pepsinogen I/II ratio 6.3 ± 1.4 (0.8 12.5) 2.7 ± 1.8 (0.3 9.6) 2.4 ± 1.6 (0.7 4.5) 5.1 ± 2.1 (0.2 17.3) [3 2402 (27.1 %) 84 (1.0 %) 4 (0.05 %) 6367 (71.9 %) B3 and [2 27 (2.2 %) 69 (5.6 %) 1 (0.08 %) 1125 (92.1 %) B2 17 (2.3 %) 114 (15.1 %) 4 (0.53 %) 620 (82.2 %) BMI 22.6 ± 3.3 (14.6 45.8) 23.5 ± 3.1 (17.4 32.6) 22.8 ± 2.2 (20.4 26.5) 23.0 ± 3.3 (13.7 48.3) \18.5 179 (27.3 %) 3 (0.5 %) 0 (0.0 %) 474 (72.3 %) C18.5 and \25.0 1777 (23.3 %) 190 (2.5 %) 7 (0.09 %) 5664 (74.2 %) C25.0 490 (19.3 %) 74 (2.9 %) 2 (0.08 %) 1974 (77.7 %) Alcohol Rarely drinking 1092 (25.9 %) 106 (2.5 %) 3 (0.07 %) 3013 (71.5 %) Usually drinking 1354 (20.5 %) 161 (2.4 %) 6 (0.09 %) 5099 (77.0 %) Smoking Lifelong nonsmoker 1557 (27.9 %) 143 (2.6 %) 2 (0.04 %) 3885 (69.5 %) Past habitual smoker 579 (18.2 %) 103 (3.2 %) 4 (0.13 %) 2498 (78.5 %) Current smoker 310 (15.0 %) 21 (1.0 %) 3 (0.15 %) 1729 (83.8 %) Gastric cancer detected Erosive 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 3 (25.0 %) Ulcerative 0 (0.0 %) 0 (0.0 %) 0 (0.0 %) 2 (16.7 %) Flat and protruded 0 (0.0 %) 0 (0.0 %) 1 (8.3 %) 0 (0.0 %) Depressed 0 (0.0 %) 1 (8.3 %) 0 (0.0 %) 5 (41.7 %) H. pylori Helicobacter pylori, BMI body mass index that the protruded type of early gastric cancer is wrongly diagnosed with a nonmalignant gastric polyp by UGI-XR. Discussion Our present results (Tables 1, 2, 3) clearly show that UGI- XR-detected gastric polyps in Japan today mostly arise from H. pylori-negative and non-atrophic mucosa with low risk of gastric cancer [14, 15]. In addition, prospective observation of the present cohort (Fig. 2) and evaluation of endoscopically diagnosed gastric polyps (Table 4) suggest that future canceration of UGI-XR-detected gastric polyps rarely occurs. As described in the Introduction, clinical steps after detecting gastric polyps by UGI-XR are still controversial in Japan. Our results indicate that it is not efficient to perform UGI-ES for all the UGI-XR-detected gastric polyps. The greatest risk of follow-up observation without UGI-ES is to erroneously diagnose protruded early gastric cancers as nonmalignant gastric polyps, but our results indicate such misdiagnoses are quite rare. Nowadays, it has been gradually accepted that UGI-XR can detect not only gastric cancer but also premalignant atrophic/hypertrophic gastritis [4, 25, 30, 32]. Our recent

320 C. Takeuchi et al. report demonstrated that UGI-XR can diagnose chronic H. pylori infection with very high accuracy (97.8 %; 1638 of 1674 subjects) [5]. Based on this accumulated evidence, we think that UGI-XR-detected gastric polyps should be evaluated together with UGI-XR-based atrophic or hypertrophic gastritis. Additional assessment of UGI-XR-based mucosal atrophy and enlarged gastric folds should help in predicting future risk of gastric cancer, because gastric polyps with malignant potential predominantly occur on H. pylori-positive gastric mucosa (Table 4). Although further data accumulation based on a longer observation period is needed, we at present conceive the follow-up observation without UGI- ES is an optimal strategy for UGI-XR-detected gastric polyps on the non-atrophic gastric mucosa. In contrast, we still recommend that UGI-XR-detected gastric polyps accompanied with atrophic or hypertrophic gastritis should be evaluated by UGI-ES as the second detailed examination. Conclusions A large-scale study of healthy people in Japan revealed that gastric polyps diagnosed by double-contrast barium X-ray radiography (UGI-XR) mostly arise on the Helicobacter pylori-negative gastric mucosa with a low risk of gastric cancer. In the prospective observation, none of the UGI-XR examinees with gastric polyps developed gastric cancer for at least the ensuing 3 years. Acknowledgments This work was supported in part by Grant-in- Aid for Scientific Research (C) from the Japan Society for the Promotion of Science (Grant Number: 25460381), in part by a grant-inaid for pioneering basic research from the Ministry of Health, Labour (H20-genome-g-006) and Welfare, and in part by a National Cancer Center Research and Development Fund (H23-A-2), Japan. Compliance with ethical standards Conflict of interest All authors declare no conflict of interest. Human rights statement and informed consent All the procedures followed were in accordance with the standards of ethics committees of the University of Tokyo on human experimentation (institutional and national) and also with the Helsinki Declaration of 1964 and later versions. Informed consent was obtained from all the study subjects. References 1. Oshima A, Hirata N, Ubukata T, Umeda K, Fujimoto I. Evaluation of a mass screening program for stomach cancer with a case-control study design. Int J Cancer. 1986;38(6):829 33. 2. Hisamichi S, Sugawara N. Mass screening for gastric cancer by X-ray examination. Jpn J Clin Oncol. 1984;14(2):211 23. 3. Fukao A, Tsubono Y, Tsuji I, Hi S, Sugahara N, Takano A. The evaluation of screening for gastric cancer in Miyagi Prefecture, Japan: a population-based case-control study. Int J Cancer. 1995;60(1):45 8. 4. Rubesin SE, Levine MS, Laufer I. Double-contrast upper gastrointestinal radiography: a pattern approach for diseases of the stomach. Radiology. 2008;246(1):33 48. 5. Yamamichi N, Hirano C, Shimamoto T, Minatsuki C, Takahashi Y, Nakayama C, Matsuda R, Fujishiro M, Konno-Shimizu M, Kato J, et al. Associated factors of atrophic gastritis diagnosed by double-contrast upper gastrointestinal barium X-ray radiography: a cross-sectional study analyzing 6,901 healthy subjects in Japan. PLoS One. 2014;9(10):e111359. 6. Cao H, Wang B, Zhang Z, Zhang H, Qu R. Distribution trends of gastric polyps: an endoscopy database analysis of 24,121 northern Chinese patients. J Gastroenterol Hepatol. 2012;27(7):1175 80. 7. Goddard AF, Badreldin R, Pritchard DM, Walker MM, Warren B. The management of gastric polyps. Gut. 2010;59(9):1270 6. 8. Carmack SW, Genta RM, Schuler CM, Saboorian MH. The current spectrum of gastric polyps: a 1-year national study of over 120,000 patients. Am J Gastroenterol. 2009;104(6):1524 32. 9. Carmack SW, Genta RM, Graham DY, Lauwers GY. Management of gastric polyps: a pathology-based guide for gastroenterologists. Nat Rev Gastroenterol Hepatol. 2009;6(6):331 41. 10. Stolte M, Sticht T, Eidt S, Ebert D, Finkenzeller G. Frequency, location, and age and sex distribution of various types of gastric polyp. Endoscopy. 1994;26(8):659 65. 11. Ljubicic N, Kujundzic M, Roic G, Banic M, Cupic H, Doko M, Zovak M. Benign epithelial gastric polyps frequency, location, and age and sex distribution. Coll Antropol. 2002;26(1):55 60. 12. Hongo M, Fujimoto K. Incidence and risk factor of fundic gland polyp and hyperplastic polyp in long-term proton pump inhibitor therapy: a prospective study in Japan. J Gastroenterol. 2010;45(6):618 24. 13. Morais DJ, Yamanaka A, Zeitune JM, Andreollo NA. Gastric polyps: a retrospective analysis of 26,000 digestive endoscopies. Arq Gastroenterol. 2007;44(1):14 7. 14. Correa P, Houghton J. Carcinogenesis of Helicobacter pylori. Gastroenterology. 2007;133(2):659 72. 15. Uemura N, Okamoto S, Yamamoto S, Matsumura N, Yamaguchi S, Yamakido M, Taniyama K, Sasaki N, Schlemper RJ. Helicobacter pylori infection and the development of gastric cancer. N Engl J Med. 2001;345(11):784 9. 16. Sipponen P, Kekki M, Siurala M. Increased risk of gastric cancer in males affects the intestinal type of cancer and is independent of age, location of the tumour and atrophic gastritis. Br J Cancer. 1988;57(3):332 6. 17. Hansen S, Wiig JN, Giercksky KE, Tretli S. Esophageal and gastric carcinoma in Norway 1958 1992: incidence time trend variability according to morphological subtypes and organ subsites. Int J Cancer. 1997;71(3):340 4. 18. Tsugane S, Sasazuki S, Kobayashi M, Sasaki S. Salt and salted food intake and subsequent risk of gastric cancer among middleaged Japanese men and women. Br J Cancer. 2004;90(1):128 34. 19. Tatematsu M, Takahashi M, Fukushima S, Hananouchi M, Shirai T. Effects in rats of sodium chloride on experimental gastric cancers induced by N-methyl-N-nitro-N-nitrosoguanidine or 4-nitroquinoline-1-oxide. J Natl Cancer Inst. 1975;55(1):101 6. 20. Sjodahl K, Lu Y, Nilsen TI, Ye W, Hveem K, Vatten L, Lagergren J. Smoking and alcohol drinking in relation to risk of gastric cancer: a population-based, prospective cohort study. Int J Cancer. 2007;120(1):128 32. 21. Koizumi Y, Tsubono Y, Nakaya N, Kuriyama S, Shibuya D, Matsuoka H, Tsuji I. Cigarette smoking and the risk of gastric cancer: a pooled analysis of two prospective studies in Japan. Int J Cancer. 2004;112(6):1049 55. 22. Gonzalez CA, Pera G, Agudo A, Palli D, Krogh V, Vineis P, Tumino R, Panico S, Berglund G, Siman H, et al. Smoking and

Gastric polyps diagnosed by double-contrast upper gastrointestinal barium X-ray radiography 321 the risk of gastric cancer in the European Prospective Investigation Into Cancer and Nutrition (EPIC). Int J Cancer. 2003;107(4):629 34. 23. Chao A, Thun MJ, Henley SJ, Jacobs EJ, McCullough ML, Calle EE. Cigarette smoking, use of other tobacco products and stomach cancer mortality in US adults: the Cancer Prevention Study II. Int J Cancer. 2002;101(4):380 9. 24. Yamamichi N, Shimamoto T, Minatsuki C, Yoshida Y, Fujishiro M, Kodashima S, Kato J, Goto O, Ono S, Niimi K, et al. Postprandial fullness correlates with rapid inflow of gastric content into duodenum but not with chronic gastritis. BMC Gastroenterol. 2011;11:140. 25. Yamamichi N, Hirano C, Ichinose M, Takahashi Y, Minatsuki C, Matsuda R, Nakayama C, Shimamoto T, Kodashima S, Ono S et al: Atrophic gastritis and enlarged gastric folds diagnosed by double-contrast upper gastrointestinal barium X-ray radiography are useful to predict future gastric cancer development based on the 3-year prospective observation. Gastric Cancer 2015. doi:10. 1007/s10120-015-0558-0. 26. Haruma K. Trend toward a reduced prevalence of Helicobacter pylori infection, chronic gastritis, and gastric cancer in Japan. Gastroenterol Clin N Am. 2000;29(3):623 31. 27. Fujisawa T, Kumagai T, Akamatsu T, Kiyosawa K, Matsunaga Y. Changes in seroepidemiological pattern of Helicobacter pylori and hepatitis A virus over the last 20 years in Japan. Am J Gastroenterol. 1999;94(8):2094 9. 28. Brown LM. Helicobacter pylori: epidemiology and routes of transmission. Epidemiol Rev. 2000;22(2):283 97. 29. Lee SY, Park HS, Yu SK, Sung IK, Jin CJ, Choe WH, Kwon SY, Lee CH, Choi KW. Decreasing prevalence of Helicobacter pylori infection: a 9-year observational study. Hepatogastroenterology. 2007;54(74):630 3. 30. Yamamichi N, Hirano C, Takahashi Y, Minatsuki C, Nakayama C, Matsuda R, Shimamoto T, Takeuchi C, Kodashima S, Ono S et al: Comparative analysis of upper gastrointestinal endoscopy, double-contrast upper gastrointestinal barium X-ray radiography, and the titer of serum anti-helicobacter pylori IgG focusing on the diagnosis of atrophic gastritis. Gastric Cancer 2015. doi:10. 1007/s10120-015-0515-y. 31. Kimura K. Chronological transition of the fundic-pyloric border determined by stepwise biopsy of the lesser and greater curvatures of the stomach. Gastroenterology. 1972;63(4):584 92. 32. Sohn J, Levine MS, Furth EE, Laufer I, Rubesin SE, Herlinger H, Lichtenstein GR. Helicobacter pylori gastritis: radiographic findings. Radiology. 1995;195(3):763 7.