800 Gut 1996; 39: 800-806 Profile of Helicobacterpylori cytotoxin derived from

two areas of Japan with different prevalence of Gut: first published as 10.1136/gut.39.6.800 on 1 December 1996. Downloaded from atrophic

S Ito, T Azuma, H Murakita, M Hirai, H Miyaji, Y Ito, Y Ohtaki, Y Yamazaki, M Kuriyama, Y Keida, Y Kohli

Abstract tritis and gastric carcinoma is not known, the Background andAim-To clarify the roles epidemiology of gastric cancers suggests that of Helicobacter pylorn cytotoxin in gastric environmental agents are important aetio- atrophy, the cytotoxin positive rate and logical factors.5 cytotoxin activity in Fukui and Okinawa, is a spiral shaped Gram where the prevalence of atrophic gastritis negative bacterium that is strongly associated and gastric cancer risk are quite different, with chronic superficial gastritis.6 As has been were studied. reported elsewhere,7 Hpylori infection in Japan Materials-Seventy three strains from has been associated with chronic atrophic Fukui and 51 from Okinawa were gastritis and intestinal metaplasia, the pre- examined. cursor lesions for gastric carcinoma.8 9 In Methods-The validation of atrQphy was addition, there is increasing evidence that per- done by endoscopy, being confirmed with sistent infection with Hpylori is a risk factor for histology. The supernatant of liquid H the development of gastric cancer, especially of pylori culture media was concentrated the distal . 10-13 20-fold, serialiy diluted, using doubling As not all persons infected with H pylori dilutions, and scored from 1 to 8. The develop gastric cancer, other cofactors must also semi-quantitated cytotoxin activity was be important. Such factors could be host related, expressed as the maximum dilution score bacterium related or related to other environ- yielding >50% A431 cell vacuolation, being mental factors. In terms of bacterium related standardised with bacterial density. factors, H pylori strains are highly diverse.`4 1' Results-The cytotoxin activity of the Thus, it is appropriate to isolate particular strains from Fukui was diverse characteristics of H strains that highly pylori might http://gut.bmj.com/ compared with that from Okinawa, affect the risk of gastric cancer development. although the cytotoxin positive rate was Recent studies have suggested that cagA, which not different. In Fukui strains, the grade is closely associated with vacuolating cytotoxin, of atrophy and the cytotoxin activity were may be an important factor for the risk of correlated (p<0.05). In addition, the cyto- developing adenocarcinoma of the stomach.'6 toxin activity of the strains from all Furthermore, the neutralising activity against patients in most of whom cytotoxin is observed more Okinawa, vacuolating on September 30, 2021 by guest. Protected copyright. showed closed-type/mild atrophy, was sig- frequently in serum samples from gastric Second Department of nificantly lower than that of the strains carcinoma patients than from healthy subjects.'7 Internal Medicine, from the patients with open-type/severe In this study, we determined the differences Fukui Medical School, Fukui, Japan atrophy in Fukui (6.46 (5.53) v 9.76 (8.80), in the presence ofvacuolating cytotoxin and its S Ito p<005), (mean (SEM)). intensity between the strains derived from the T Azuma Conclusion-The difference in profile of patients in Fukui and Okinawa prefectures, in H Murakita M Hirai the cytotoxin activity in the two areas was which the prevalence of atrophic gastritis and H Miyaji related to the difference in the prevalance the gastric cancer risk were quite different, to Y Ito ofatrophic gastritis. clarify their roles in the development of gastric Y Ohtaki Y Yamazaki (Gut 1996; 39: 800-806) atrophy or the increase in gastric cancer risk, M Kuriyama or both. Keywords: Helicobacterpylori, vacuolating cytotoxin, Division of Internal gastric atrophy. Medicine, Okinawa Chubu Hospital, Methods Okinawa, Japan Y Keida Although the incidence and the death rate of Patients Division ofInternal gastric adenocarcinoma in Japan is high com- The subjects were new patients who were Medicine, pared with that in other developed countries,l examined for H pylori status by endoscopy in Aiseikai-Yamashina in the death Hospital, Kyoto, Japan large intracountry differences rates the Second Department of Internal Medicine, Y Kohli of gastric cancer have been reported.2 3 The Fukui Medical School in Fukui, the middle Correspondence to: death rate in Okinawa prefecture is lower than part of Japan, between February and Dr S Ito, that in any other prefecture in Japan, including November 1994 and Okinawa Chubu Hospital 2nd Department of Intemal Medicine, Fukui Medical Fukui prefecture. In addition, the prevalence in Okinawa, the most south west part ofJapan, School, 23-3, Shimoaizuki, of atrophic gastritis in Okinawa is also signifi- between September and December 1994. Matsuoka-cho, Yoshida-gun, Fukui, 910-1 1, Japan. cantly lower than that in any other prefecture However, nine of 123 examined patients in Accepted for publication in Japan.3 4 Although the precise mechanism Fukui and eight of 106 in Okinawa had various 8 July 1996 underlying the development of atrophic gas- diseases other than gastritis or peptic ulcer, that Hpylori cytotoxin and atrophic gastritis 801

is, two with gastric , three with gastric above using a sterilised endoscope, put into cancer, and four with remnant stomach in Cary-Blair-N transport medium (Nissui Fukui and four with gastric polyp, one with Seiyaku Co. Tokyo, Japan), and cultured on gastric cancer, and three with remnant stomach Trypticase Soy Agar-II with 5% sheep blood Gut: first published as 10.1136/gut.39.6.800 on 1 December 1996. Downloaded from in Okinawa, and were excluded. The subjects (TSA-II) plate (Nippon Bekton Dickinson, had not been receiving any antisecretory agents Tokyo, Japan) in the microaerophilic condition or anti-H pylori treatment in the long term (02: 5%, CG2: 15%, N2: 80%) for five days. before this study. Seventy three of 114 subjects The specimens obtained from Okinawa pre- in Fukui (male/female: 45/28) and 51 of 98 fecture were sent by air in the same transport subjects in Okinawa (male/female: 24/27) were medium in an ice box and cultured within 24 H pylori positive. In Fukui, 41 patients had hours after biopsy. In addition, a histological mild epigastralgia but no gross findings or study with Giemsa staining was done on the diseases in their stomach, that is, had only type samples from Okinawa prefecture and con- B gastritis including erosive or superficial gas- firmed by successful isolation ofculture regard- tritis, or both, six with gastric ulcer, and 26 less of transport by air. H pylori was identified with duodenal ulcer. Similarly in Okinawa, 33 as Gram negative bacilli with urease, catalase, had gastritis, nine gastric ulcer, and nine duo- and oxidase activity. A few colonies ofthe third denal ulcer. The duodenal and gastric ulcer cultured H pylori isolates (approximately 105 patients also had gastritis. The diagnosis was cfu/ml) on a TSA-II plate, which had been well confirmed endoscopically by two endoscopists. adapted in the in vitro environment, were then The mean (SD) age of the patients in Fukui cultured in 20 ml of Brucella broth supple- and Okinawa was 55.5 (15i4) (range, 16-87) mented with 10% fetal calf serum. Broth years old and 61-0 (12X1) (range, 35-88) years cultures were incubated with a gyratory shaker old, respectively. Although the percentage of at 120 rpm in a microaerophilic condition for duodenal ulcer patients in Fukui seemed to be three days. The growth ofHpylori was assessed high, the differences in patients found to have by comparison of the optical density at 560 nm ulcers (43.8% in Fukui v 35.3% in Okinawa), of the broth culture medium of each H pylorn the degree of symptoms in the two prefectures isolate with that of the culture medium alone. was not significantly different.

Validation ofatrophy Firstly, histological atrophy was examined in 84 of 124 biopsied samples of the antrum with haematoxylin and eosin staining from both

prefectures (33 from Fukui and 51 from http://gut.bmj.com/ Okinawa) according to the Sydney system,'8 scored 0; none, 1; mild, 2; moderate and 3; severe and compared with endoscopically diag- nosed atrophy as described below. Secondly, in all cases, with the location of the endoscopic atrophic border identified by endoscopy, the

grade of atrophy was diagnosed according to on September 30, 2021 by guest. Protected copyright. the criteria of Kimura and Takemoto as re- ported previously (Fig 1).9 19-21 The endoscopic atrophic border marks the furthest front of the extent of atrophic gastritis from the antrum to the body in the stomach as a result of chronic type B gastritis, although gastritis may be localised to the antrum in Western patients.2123 In brief, by identifying the location of the endoscopic atrophic border in the stomach, the grade of atrophy was divided into two types, closed-type/mild extension of atrophy and open-type/severely extended atrophy. If the border was on the lesser curvature of the stomach, it was defined as closed-type Figure 1: The extension ofendoscopic atrophic border and were subdivided into patterns ofendoscopically diagnosed atrophic gastritis by (C-type). They Cl, C2, Kimura and Takemoto. The endoscopic atrophic border and C3 patterns. Ifthe border was shifted orally marks thefurthestfront ofthe extent ofatrophic gastritis in and did not exist on the lesser curvature, it was the stomach as a result ofchronic type B gastritis in defined as open-type (0-type), and subdivided J7apanese patients. J'apanese patients tended to have large areas ofatrophy and intestinal metaplasia in the stomach. into 01, 02, and 03 patterns. Therefore, the By identifying the location ofthe endoscopic atrophic border gastric atrophy was graded as Cl, C2, C3, 01, in the stomach, the grade ofatrophy was divided into two 02, and 03 according to their severity (Fig 1). types, closed-type/mild extension ofatrophy and open-type! severely extended atrophy. Ifthe border was on the lesser curvature of the stomach, it was defined as closed-type (C-type). They were subdivided into Cl, C2, and C3 Preparation ofthe strains patterns. If the border was shifted orally and did not exist on the lesser curvature, it was defined as open-type (0-type), Gastric biopsy samples were taken from the and subdivided into 01, 02, and 03 patterns. Therefore, lesser curvature of the antrum and large the gastric atrophy was graded as Cl, C2, C3, 01, 02, curvature of the body of all patients described and 03 according to their severity. 802 Ito, Azuma, Murakita, Hirai, Miyaji, Ito, Ohtaki, Yamazaki, Kuriyama, Keida, Kohli

Evaluation ofcytotoxin production and performed on those from age 40 to 79. The quantification ofactivity cytotoxin positive rate and the intensity were A431 (ATCC CRL 1555) cells were derived determined for every decade in Fukui and from human squamous cell carcinoma and Okinawa. kindly provided by Dr Shimizu (Keio Uni- Gut: first published as 10.1136/gut.39.6.800 on 1 December 1996. Downloaded from versity, Tokyo, Japan). Cells were maintained as described previously.24 A431 cells (0.2 x 1i0 Statistics cells) were inoculated into each well of a 96 The cytotoxin positive rate was analysed by the well microtitre plate. After a 24 hour incu- X2 test. The relation between atrophy diag- bation period, the medium was replaced with nosed histologically and endoscopically and 90 ,u of fresh medium supplemented with 10 that between the grade of gastric atrophy and mmol/l ammonium chloride to potentiate cyto- the degree of cytotoxin activity were analysed toxin activity. The supernatant of H pylori by the Spearman rank correlation test. The culture was concentrated 20-fold by using a difference in cytotoxin activity between Fukui 100 kDa ultrafiltration membrane (Ultrafree and Okinawa was compared by the Student's C3HK, Millipore, Maryland, USA) and then t test. sterilised by passage through a 0.22 ,um filter (Ultrafree C3GV, Millipore, Maryland, USA). For determination of the relative cytotoxin Results activity, the concentrated culture supernatants There was no significant difference in the cyto- were serially diluted, using doubling dilutions toxin positive rate between Fukui and Okinawa of 1, 2, 4, 8, 16, 32, 64, and 128 by the A431 for any of the strains isolated (Table I), or culture medium and scored from 1 to 8. between age groups (Table II), although the Ten [1l of aliquot was added to 90 pI of A431 number in some age groups was too small for medium in a 96 well plate. After a 24 hour comparison. incubation period, cells were examined micro- The histological atrophy of the antrum and scopically for the presence of intracellular the endoscopically diagnosed atrophy were vacuolisation. The relative cytotoxin activity positively correlated according to the Spear- was expressed as the maximum dilution score man rank correlation test (Fig 2) (p<0-01). yielding >50% A431 cell vacuolation. The The semi-quantitated cytotoxin activity of relative activity was defined as 0.5 in a sample the strains from Fukui was highly diverse com- of score 1 supernatant that showed vacuolation pared with that of Okinawa (Fig 3). In Fukui in less than 50% ofthe cells. For standardisation strains, there was a positive correlation be- by bacterial number, the semi-quantitated cyto- tween the grade of endoscopically diagnosed toxin activity was expressed by dividing the gastric atrophy and the cytotoxin activity activity with each remainder of optical relative http://gut.bmj.com/ density between a 3 day old liquid culture of H pylori and culture medium alone. We compared the positive rate of H pylori cytotoxin and their intensity in all patients and 03 F- in the patients with gastritis alone between two prefectures. Then, an age matched study was

02 F- a I I on September 30, 2021 by guest. Protected copyright. TABLE I Examinedpatients in Fukui and Okinawa and their cytotoxin positive rate ulcer Duodenal uker Total Gastritis Gastric 01 F- S I Examined cases in Fukui 78 8 28 114 a, Hpylori positive cases 41 6 26 73 (0 Cytotoxin positive cases 31 5 14 50 rate 53-8 57.5 (D I Cytotoxin positive (%) 75-6 83-3 C3 F- l Examined cases in Okinawa 77 10 1 1 98 AI Hpylori positive cases 33 9 9 51 Cytotoxin positive cases 24 4 6 34 Cytotoxin positive rate (%) 72-7 44-4 66-7 66-7 C2 [- l S a in Fukui and Okinawa was not different. In each group, I The cytotoxin positive rate of the strains the rate was slightly higher in the strains from atrophic gastritis than from duodenal ulcer. Gastritis patients had mild epigastralgia but no gross findings or diseases in their stomach, that is, had only type B gastritis including erosive or superficial gastritis, or both. Cl h 1

TABLE ii Cytotoxin positive rate according the decadefrom age 40 to 79 in Fukui and . Okinawa 0 1 2 3 Age 40-49 50-59 60-69 70-79 Total None Mild Moderate Severe Fukui Score Hpylori positive cases 13 18 17 1 1 59 the Cytotoxin positive cases 10 13 11 8 42 Figure 2: Correlation between histological atrophy of Cytotoxin positive rate (/) 76-9 72-2 647 72.7 71-2 antrum and endoscopically diagnosed atrophy. They had a positive correlation (p<0.01). The x axis is histological Okinawa the axis is Hpylori positive cases 6 10 20 10 46 atrophy according to the Sydney system and y Cytotoxin positive cases 3 8 13 7 31 endoscopically diagnosed atrophy according to the criteria of Cytotoxin positive rate (%) 50.0 80-0 65-0 70.0 67-4 Kimura and Takemoto. The histological atrophy was scored as 0; none, 1; mild, 2; moderate and 3; severe. was graded as The cytotoxin positive rate of the strains in Fukui and Okinawa was not affected by age from 40 Endoscopically diagnosed atrophy Cl, C2, to 79. C3, 01, 02, and 03 according to their severity. Hpylori cytotoxin and atrophic gastritis 803

A When age matched for every decade, a 30 F 30 r similar finding was clearly observed for those over 60 years old (Figs 4 and 5). The cytotoxin

activity was more diverse in Fukui than in Gut: first published as 10.1136/gut.39.6.800 on 1 December 1996. Downloaded from . Okinawa, correlated with the grade of atrophy. 0 The cytotoxin activity from the patients with 0 open-type/severe atrophy in Fukui was higher 20K S 20 F- than that of all strains in Okinawa. ._2 0 0 0 a 0 0 Discussion ._Co L 0) In this study, the cytotoxin activity of H pylori 0 0 0 0 0 0 0 * a strains isolated from Fukui was highly diverse 0 0 compared with that of the strains from C.. . a 0 0 10 H 10 - Okinawa. The intensity of the cytotoxin S 0 . activity in Fukui had a positive correlation with .3a 0 09 the degree of gastric atrophy when examined in all 73 cases and in 41 cases of gastritis alone 0 (p<0 05), especially in those over 60 years old. AIa*C1C2C3010203 0 More virulent strains on the point of cytotoxin 0~~~~~ 0 production were associated with the patients in Cl C2' C3 01 0'2 03 Cl C2 03 01 02 03 Fukui who had open-type/severe atrophy. In contrast, the virulency of the strains from B Okinawa was weak and the degree of atrophy 30 r 30 r was also mild. In several studies, cytotoxin positive strains have been more associated with peptic ulcers, chronic gastritis with severe and eventually atrophic gastritis.2527 However, the cytotoxin positive 0 rate was not different between Okinawa and 20k- 20k- Fukui, regardless of the differences in gastric atrophy. This suggested that the intensity of '._0 0 0 cytotoxin activity was more closely related to C. 0 cc 0 the development of gastric atrophy. 0 0 Our study showed that endoscopically diag- 0 * 0 0 x~ nosed atrophy was correlated with the histo- http://gut.bmj.com/ 0 0 0 0 0 0 0 logical atrophy of the gastric antrum. In many 10 - 10 _ 0 0 0 previous works, the endoscopically diagnosed S 0 0 is reported to be well correlated with a 0 atrophy S the extension of pyloric gland atrophy and intestinal metaplasia and also negatively corre- 0 lated with the acid output of the stomach.'9-22 0 a3 0 0

0 As mentioned above, the endoscopic atrophic on September 30, 2021 by guest. Protected copyright. 0 0 border marks the furthest front of the extent of C'l 02 03 01 02 03 Cl C2 C3 01 02 03 atrophic gastritis in the stomach as a result of Figure 3: Correlation between the grade ofgastric atrophy and the intensity ofcytotoxin chronic type B gastritis in Japanse patients. activity ofisolated H pylon (A) in allpatients examined and (B) in the patients with Japanese patients tended to have a large area gastritis alone in Fukui (0) and Okinawa (0). The cytotoxin activity ofisolated H pylon of atrophy and intestinal metaplasia in the was semi-quantitatively described. The grade ofatrophy was expressed as Cl, C2, C3, 01, 02, and 03 according to its severity. The cytotoxin activity was correlated with the stomach, although gastritis may be localised to endoscopically diagnosed gastric atrophy in allpatients (A) and in gastritis alone (B) in the antrum and there is much more multi-focal Fukui (p<0 05). Most ofthe patients in Okinawa had closed-type/mild atrophy. The atrophic gastritis in the West.2'23 28 The cytotoxin activity (mean (SEM)) ofall the strainsfrom Okinawa was significantly lower than thatfrom the patients with open-type/severe atrophy in Fukui (6.46 (5.53) v 9-76 Sydney system for the classification of gastritis (8.80), p0 05). atrophy or intestinal metaplasia, or both, 804 Ito, Azuma, Murakita, Hirai, Miyaji, Ito, Ohtaki, Yamazaki, Kuriyama, Keida, Kohli

A more factors and multiple steps are associated 307 307 with the development of gastric adeno- carcinoma.3' Recent studies have shown that cytotoxin production does not necessarily require the cagA gene.32 33 However, cytotoxin Gut: first published as 10.1136/gut.39.6.800 on 1 December 1996. Downloaded from production is associated with cagA expres- sion.26 27 CagA positive strains were more 20 F- 20 H virulent and more immunologically strong.27 In addition, by the serological studies, the anti- 0 S CagA antibody or neutralising antibody to cytotoxin in their sera were more associated 0 with the risk for gastric cancer.'6 17 As environ- 0 mental exposure early in life has been impli- 10 e 10l- cated in the pathogenesis ofgastric cancer,5 the 0 exposure to more virulent strains may be 0 0 0 correlated with the development of gastric 0 adenocarcinomas. The large difference in the . prevalence of atrophic gastritis between Fukui 0 and Okinawa may depend on the difference in the profile of cytotoxin activity of Hpylori and 01 C2 C3 01 02 03 Cl C2 C3 01 02 03 may eventually reflect the difference in gastric cancer risk to some degree. Although the simultaneous gastric infection B with two H pylori strains has been reported at 30r 30 r frequencies of 10-13%/l5 the PCR-RFLP patterns using the ureB locus showed that almost all patients had the same strain 0 throughout their lifetime.34 Megraud et a135 reported by the analysis of molecular finger- printing of H pylori that differences were not 20 H 20 H due to a change of the strain but to the minor genomic change within the same strain during C.) 0 the long H pylori infection. Therefore, the 0 ._ 0 0 virulency of the strain was continuously con- 0 0 0 served for a long period, albeit with some 0 exceptions.36 In addition, divergence ofgenetic x 0 10; 10 sequences for the vacuolating cytotoxin among http://gut.bmj.com/ 0* 0 8 H pylori strains has been reported and the * 0 0 variation in the length of the vacA gene could reflect the phenotype in the strains.37 38 Recently, differences in the cytotoxin activity 0 have been also reported and genetically

0 0 determined.39 In addition, cytotoxin activity is C1 C2 C3 01 02 03 C1 C2 C3 01 02 03 also affected by their secretory function, which on September 30, 2021 by guest. Protected copyright. Figure 4: Correlation between the grade ofgastric atrophy and the cytotoxin activity of has been reported to be similar with the IgA patients in their 40s (A) and 50s (B) in Fukui (0) and Okinawa (0). The grade of protease type of exoprotein.40 Further studies atrophy was expressed as Cl, C2, C3, 01, 02, and 03 according to its severity. Although the cytotoxin activity ofthe strainsfrom the patients in their 40s showed no difference are necessary to clarify the mechanism to between Fukui and Okinawa, thatfrom the patients in their 50s in Fukui was diverse and induce different cytotoxin activity. some ofthem who had open-type/severe atrophy showed high cytotoxin activity. H pylori infection is established in child- hood.4' 42 Several epidemiological studies suggested that the incidence of H pylori although the grade of their atrophy was closed- infection increases in adolescence in Japan and type/mild. Gastric cancer patients also had other developed countries.43 44 Therefore, duodenal ulcers in about 1.2% of the resected patients over 60 years old may have been specimens in Japan.29 Because, in Japan, exposed and affected with Hpylori infection for almost all patients who had been infected with more than 20 years. The association between H pylori may have a risk for the development the period of H pylori infection and the de- of atrophy, we compared the gastric atrophy in velopment of gastric atrophy has been all patients in addition to the patients with suggested." 12 41 45 In this study, a positive gastritis alone. Extension of atrophic gastritis correlation was seen between gastric atrophy from the antrum to the corpus probably and cytotoxin activity especially in those over depends on a number of genetic and environ- 60 years old. Therefore, longer exposure to H mental factors, including diet, alcohol con- pyloi infection as well as the exposure to a sumption, bile reflux, and infection with H more virulent factor may be necessary for the pylori. Of them, infection with H pylori as well development of severe gastric atrophy. Under as the intensity of cytotoxin is thought to be a 60 years old, the difference between Fukui and powerful trigger of the extension of gastric Okinawa was not so clear, although the atrophy. examined number of patients under 50 years Risk factors for atrophic gastritis are also old was too small to make a comparison. The linked to those for gastric cancer,4 30 although other reason may be that the strain from the Hpylori cytotoxin and atrophic gastritis 805

A not only the prevalence of H pylori infection 30 30 and the analysis of cytotoxin positive rate but also the analysis of the virulency in cytotoxin . activity will be needed to clarify the association between H pylon infection and gastric cancer Gut: first published as 10.1136/gut.39.6.800 on 1 December 1996. Downloaded from risk. In contrast, the fact that the strains without 120 - *20 _ cytotoxin activity or with lower cytotoxin 0~~~~~ activity in Fukui were also found in patients Xo with severe atrophic gastritis showed that factors others than H pylori cytotoxin might also be responsible for atrophic gastritis. Many Q10_ * 10 0 o pathogenic factors ofHpylori strains have been 0 reported.6 Differences in the host or in the host 0~~~~~~~~~~~~~~ 0 ~ ~ ~ 00 strain relation have also been suggested as a cause of the difference in the prevalence of atrophic gastritis and the death rate of gastric cancer between Fukui and Okinawa.50 51 Food and life style may also affect the development C1l C2 C3 01(02 03 Cl C2 C3 01 02 03 ofgastric carcinoma and the difference in death rate.3 31 The HLA-DQA gene is a candidate B gene for the host's susceptibility in the estab- 30 - 30 0 0 lishment of Hpylori infection and the develop- ment of gastric atrophy.52 ` The Leb+ antigen 0~~~~~~~~~~~~~~ is one of the factors for the attachment of H pylori infection.54 We are now studying these factors in Fukui and Okinawa. In conclusion, the cytotoxin activity ob- > 20 0 20 served in Fukui strains was highly diverse and 0 0 was correlated with the development of gastric 0~~~~~~~~~ atrophy. The difference in profile of the cyto- toxin activity in Fukui and Okinawa was 0~~~~~~~~~ related to the difference in the prevalence of C10 10 _ o atrophic gastritis, which eventually may reflect 0~~~~~~~~~~~~~~ the difference in the gastric cancer risk between

them. http://gut.bmj.com/

1 Watanabe S, Arimoto H. Long term trends in cancer 0 I1 mortality rates and cumulative rates from 1955 to 1987 C0 C2 C3 021003 Cl C2 C3 01 02 03 inJapan (II).JpnJClin Oncol 1989; 19: 413-25. 2 Statistics and Information Department, Minister's Figure S: Correlation between the grade ofgastric atrophy and the cytotoxin activity of Secretariat. Vital Statistics of J7apan. Tokyo: Ministry of patients who were in their 60s (A) and 70s (B) in Fukui () and Okinawa (O). The Health and Welfare, 1992 (in Japanese). grade ofatrophy was expressed as Cl, C2, C3, Ol, 02, and 03 according to its severity. 3 Tsugane S, Kabuto M, Imai H, Gey F, Tei Y, Hanaoka T, on September 30, 2021 by guest. Protected copyright. The the strains the in their 60s in Fukui was et al. Helicobacter pylori, dietary factors and atrophic cytotoxin activity of from patients highly gastritis in five Japanese populations with different gastric diverse and correlated significantly with the grade ofatrophy (p<0 05). On the contrary, cancer mortality. Cancer Causes Control 1993; 4: thatfrom Okinawa showed relatively low activity. 297-305. 4 Kabuto M, Imai H, Tsugane S, Watanabe S. Correlation between atrophic gastritis prevalence and gastric cancer mortality among middle-aged men in 5 areas in Japan. J Epidemiol 1993; 3: 35-9. youngerlogcasudessugetedageneration in Okinawaoreltinis changingete0 5 Howson CP, Hiyama T, Wynder EL. The decline in gastric 0 andh likelyrvlneoto have a yoiifcinadtersstrong cytotoxin activity. cancer: epidemiology of an unplanned triumph. Epidemiol Rev 1986; 8: 1-27. A oflongerasriccacechronological1>3 6 study4 Hweeron a youngerevr0 6 Blaser MJ. Helicobacterpylori and the pathogenesis ofgastro- generation will be needed to clarify this duodenal inflammation. J Infect Dis 1990; 161: 626-33. 7 Craanen ME, Dekker W, Blok P, Ferwerda J, Tytgat GNJ. problem. Intestinal metaplasia and Helicobacter pylori: an In Okinawa prefecture, the prevalence of H endoscopic bioptic study of the gastric antrum. Gut 1992; Figure5:Corre ationbetweentegaeo gastricatopyandtercyooink activit pyof i 33: 16-20. patensho er i teir60 ()pylorindect70sinfection(B) Vainoukuhas been(0) andreportedorkinawal(0).Thto be lower 8 Sakaki N, Momma K, Yamada Y, Tadokoro Y, Tajima T. grade ~ ~ ofeomno ~ thatixrse thansC,C,C,0,0,ad0codnwaatrophythat in other prefectures in Japan.'oritgsevrityWe Helicobacter pylon infection and the development of atrophic gastritis assessed by endoscopy. Eur Jf Gastro- Thecyotxinacivtyfhestainscfromalso obtainedthe patienta Insimilartheirm6sfindingin Fukyoiinwasehighlythis study, enterol Hepatol 1992; 4: 585-7. although the difference was not significant 9 Kohli Y, Kato T, Iwaki M, Ito S, Suzuki K. Endoscopic diagnosis of Helicobacter pylori distribution in gastric (52-0% v 64.0%. p>0 05). The prevalence of mucosa of patients with chronic gastritis. Eur _J Gastro- atrophic gastritis and the death rate of gastric enterol Hepatol 1993; 5 (suppl 1): S127-3 1. 10 Blaser MJ, Kobayashi K, Cover TL, Cao TL, Feurer ID, cancer were also significantly lower than in Perez-Perez GI. Helicobacter pylori infection in Japanese other prefectures in japan.2- Many epidemio- patients with adenocarcinoma ofthe stomach. IntJ Cancer 1993; S5: 799-802. 11 Nomura A, Stemmerman GN, Chyou PH, Kato I, Perez- Perez GI, Blaser MJ. Helicobacter pylori infection and gastric carcinoma in a population of Japanese-Americans in Hawaii. NEnglJMed 1991; 325: 1132-6. 12 Parsonnet J, Friedman GD, Vandersteen DP, Chang Y, Vogelman JH, Orentreich N, et al. Helicobacter pylori infection and the risk of gastric carcinoma. N EnglJf Med 1991; 325: 1127-31. 13 The Eurogast Study Group. An international association between Helicobacter pylon infection and gastric cancer. Lancet 1993; 341: 1359-62. 806 Ito, Azuma, Murakita, Hirai, Miyaji, Ito, Ohtaki, Yamazaki, Kuriyama, Keida, Kohli

14 Akopyanz N, Bukanov NO, Westblom TU, Kresovich S, pylon) by southern blotting with a urease gene proble. Jf Berg DE. DNA diversity among clinical isolates of Gastroenterol 1994; 29: 120-4. Helicobacter-pylori detected by PCR-based RAPD 35 MegraudF, CantetF,HuaJ, Birac C, CaylaR,LamouliatteH. fingerprinting. Nucleic Acids Res 1992; 20: 5137-42. Molecular fingerprinting of Helicobacter pylon strains 15 Prewett EJ, Bickley J, Owen RJ, Pounder RE. DNA patterns shows a low intrapatient heterogeneity.

of Helicobacter pylorn isolated from gastric antrum, body, 1994; 106 (suppl): A136. Gut: first published as 10.1136/gut.39.6.800 on 1 December 1996. Downloaded from and duodenum. Gastroenterology 1992; 102: 829-33. 36 Taylor NS, FoxJG, Akopyants NS, Berg DE, Thompson N, 16 Blaser MJ, Perez-Perez GI, Kleanthous H, Cover TL, Shames B, et al. Long-term colonization with single and Peek RM, Chou PH, et al. Infection with Helicobacter multiple strains of Helicobacter pylon assessed by DNA pylon strains possessing cagA is associated with an fingerprinting. JClinMicrobiol 1995; 33: 918-23. increased risk of developing adenocarcinoma of the 37 CoverTLTummuruMKR,CaoP,ThompsonSA,BlaserMJ. stomach. Cancer Res 1995; 55: 2111-5. Divergence of genetic sequences for the vacuolating 17 Hirai M, Azuma T, Ito S, Kato T, Kohli Y, Fujiki N. High cytotoxin among Helicobacter pylon strains. Jf Biol Chem prevalence of neutralizing activity to Helicobacter pylori 1994; 269: 10566-73. cytotoxin in serum of gastric-carcinoma patients. Int Jf 38 Phadnis SH, Ilver D, Janzon L, Normark S, Westblom TU. Cancer 1994; 56: 56-60. Pathological significance and molecular characterization 18 Price AB. The Sydney System: histological division. Jf of the vacuolating toxin gene of Helicobacter pylon. Infect Gastroenterol Hepatol 1991; 6: 209-22. Immun 1994; 62: 1557-65. 19 Kimura K, Takemoto T. An endoscopic recognition of the 39 Atherton JC, Cao P, Peek RM, Tummuru MKR, Blaser MJ, atrophic border and its significance in chronic gastritis. Cover TL. Mosaicism in vacuolating cytotoxin alleles of Endoscopy 1969; 3: 87-97. Helicobacterpylon. Jf Biol Chem 1995; 30: 17771-7. 20 Kimura K. Chronological transition of the fundic-pyloric 40 Schmitt W, Haas R. Genetic analysis of the Helicobacter border determined by stepwise biopsy of the lesser and pylori vacuolating cytotoxin: structural similarities with greater curvatures of the stomach. Gastroenterology 1972; the IgA protease type of exported protein. Mol Microbiol 63: 584-92. 1994; 12: 307-19. 21 Kimura K, Satoh K, Ido K, Taniguchi Y, Takimoto T, 41 Blaser MJ, Chyou PH, Nomura A. Age at establishment of Takemoto T. Gastritis in the Japanese stomach. Scand J Helicobacter pylori infection and gastric carcinoma, gastric Gastroenterol 1996; 31 (suppl 214): S17-20. ulcer, and duodenal ulcer risk. Cancer Res 1995; 55: 22 Kohli Y, Pfeiffer CJ, Kutty KP, BarrowmanJA, Heughan C, 562-5. Kepkay DL. Endoscopic diagnosis of intestinal meta- 42 Mendall MA, Goggin PM, Molineaux N, Levy J, Toosy T, plasia in Canada and Japan. Jf Clin Gastroenterol 1981; 3 Strachan D, et al. Childhood living conditions and (suppl 1): S29-33. Helicobacter pylon seropositivity in adult life. Lancet 1992; 23 Correa P. Chronic gastritis: A clinico-pathological 339: 896-7. classification. Am Jf Gastroenterol 1998; 83: 504-9. 43 Asaka M, Kimura T, Kudo M, Takeda H, Mitani S, 24 Murakita H, Hirai M, Ito S, Azuma T, Kato T, Kohli Y. Miyazaki T, et al. Relationship of Helicobacter pylori to Vacuolating cytotoxin production by Helicobacter pylori serum pepsinogens in an asymptomatic Japanese popu- isolates from peptic ulcer, atrophic gastritis and gastric lation. Gastroenterology 1992; 102: 760-6. carcinoma patients. Eur J Gastroenterol Hepatol 1994; 6 44 Graham DY, Malaty HM, Evans DG, Evans DJ, Klein PD, (suppl 1): S29-31. Adam E. Epidemiology of Helicobacter pylori in an 25 Fox JG, Correa P, Taylor NS, Thompson N, Fontham E, asymptomatic population in the United States. Janney F, et al. High prevalence and persistence of Gastroenterology 1991; 100: 1495-501. cytotoxin-positive Helicobacter pylori strains in a popu- 45 Forman D, Webb P, Parsonnet J. Helicobacter pylon and lation with high prevalence of atrophic gastritis. Am J gastric cancer. Lancet 1994; 343: 243-4. Gastroenterol 1992; 87: 1554-9. 46 Fukuda H, Saito D, Hayashi S, Hisai H, Ono H, Yoshida S, 26 Cover TL, Cao P, Lind CD, Tham KT, Blaser MJ. et al. Helicobacter pylori infection, serum pepsinogen level Correlation between vacuolating cytotoxin production by and gastric cancer: A case-control study in Japan. Jpn Jf Helicobacter pylon isolates in vitro and in vivo. Infect Cancer Res 1995; 86: 64-71. Immun 1993; 61: 5008-12. 47 Ching CK, Lam SK. Helicobacter pylon epidemiology in 27 Covacci A, Censini S, Bugnoli M, Petracca R, Burroni D, relation to peptic ulcer and gastric cancer in South and Macchia G, et al. Molecular characterization of the North China. Jf Gastroenterol Hepatol 1994; 9: S4-7. 128-kDa immunodominant antigen of Helicobacter pylori 48 Sierra R, Mufhoz N, Pefia AS, Biemond I, Duijn W, associated with cytotoxicity and duodenal ulcer. Proc Natl Lamers CBHW, et al. Antibodies to Helicobacterpylon and Acad Sci 1993; 90: 5791-5. pepsinogen levels in children from Costa Rica: com- 28 Kohli Kato Suzuki Tada N. Incidence of two areas with different risks for stomach Y, T, K, T, Fujiki parison http://gut.bmj.com/ of atrophic gastritis with age in Japan and Canada. J3pn cancer. Cancer Epidemiol Biomarkers Prev 1992; 1: JMed 1987; 26: 158-61. 449-54. 29 Hosokawa 0. Clinical study of 2157 cases with early gastric 49 Palli D, Decarii A, Cipriani F, Sitas F, Forman D, Amdori D, cancer in the Division ofSurgery, Fukui Prefectural Hospital. et al. Helicobacter pylon antibodies in areas of Italy at Fukui, Yoshidakinbundo, 1996 (in Japanese). varying gastric cancer risk. Cancer Epidemiol Biomarkers 30 Sipponen P, Kekki M, Haapakoski J, Ihamaki T, Siurala M. Prev 1993; 2: 37-40. Gastric cancer risk in chronic atrophic gastritis: statistical 50 CrabtreeJE, TaylorJD, WyattJI, Heatly RV, Shallcross TM, calculations ofcross-sectional data. IntJ Cancer 1985; 35: Tompkins DS, et al. Mucosal IgA recognition of 173-7. Helicobacterpylon 120 kDa protein, peptic ulceration, and 31 Correa P. Human gastric carcinogenesis: a multistep and gastric pathology. Lancet 1991; 338: 332-5. multifactorial process-First American Cancer Society 51 Schlemper RJ, Werf SDJ, Vandenbroucke JP, Biemond I, on September 30, 2021 by guest. Protected copyright. Award lecture on cancer epidemiology and prevention. Lamers CBHW. Seroepidemiology ofgastritis in Japanese Cancer Res 1992; 52: 6735-40. and Dutch working populations: evidence for the 32 Tummuru MKR, Cover TL, Blaser MJ. Mutation of the development of atrophic gastritis that is not related to cytotoxin-associated cagA gene does not affect the Helicobacterpylori. Gut 1995; 37: 199-204. vacuolating cytotoxin activity of Helicobacter pylori. Infect 52 Azuma T, Konishi J, Tanaka Y, Hirai M, Ito S, Kato T, Immun 1994; 62: 2609-13. et al. Contribution of HLA-DQA gene to host's response 33 Xiang Z, Censini S, Bayeli PF, Telford JL, Figura N, against Helicobacterpylon. Lancet 1994; 343: 542-3. Rappuoli R, et al. Analysis of expression of CagA and 53 Beales ILP, Davey NJ, Pusey CD, Lechier RI, Calam J. VacA virulence factors in 43 strains of Helicobacter pylon Long-term sequelae of Helicobacter pylon gastritis. Lancet reveals that clinical isolates can be divided into two major 1995; 346: 381-2. types and that CagA is not necessary for expression of the 54 Boren T, Falk P, Roth KA, Larson G, Normark S. Attach- vacuolating cytotoxin. Infect Immun 1995; 63: 94-8. ment of Helicobacter pylon to human gastric epithelium 34 TonokatsuY,HayashiT,FukudaY,TamuraT,ShimoyamaT. mediated by blood group antigens. Science 1993; 262: A clinico-epidemiological analysis ofHelicobacterpylori (H 1892-5.