- Email: [email protected]
GASTRITIS AND GASTRIC CANCER
HELICOBACTER PYLORI, PART I
0889-8553 / 00 $15.00
+ .OO
GASTRITIS AND GASTRIC CANCER Asia Ken Kimura, MD
Japan has the highest annual incidence of gastric cancer in the world. The prevalences of Helicobacter pylori infection and atrophic gastritis (AG) in Japan are also high. H. pylori infection causes chronic gastritis, and it leads to development and progression of AG and intestinal metaplasia, which are considered precursor lesions of gastric cancer. This article describes the relationship of H. pylori infection and AG and gastric cancer in Japan. PREVALENCE OF HELICOBACTER PYLORI INFECTION AND ATROPHIC GASTRITIS
In 1992, Asaka et a12 reported the prevalence of H. pylori infection in the asymptomatic Japanese population using serum anti-H. pylori IgG antibody. The prevalence was less than 10% in persons aged 0 to 9 years, was about 20% in persons aged 10 to 19 years, increased with age, and reached the plateau of 80% in persons aged 40 to 49 years. The prevalence in persons born after 1950 increased at approximately 1%per year, and the prevalence in persons born before 1949 was high, relatively constant, and approximately double of that of the 30- to 40-year age group. There are two cohorts that differ in the risk for H. pylori in Japan. In 1993, Fukao et a17 studied serum samples of 1815 healthy blood donors and reported the prevalence of H. pylori infection and presumed
From the Department of Gastroenterology, Jichi Medical School, Tochigi, Japan
GASTROENTEROLOGY CLINICS OF NORTH AMERICA
-
VOLUME 29 NUMBER 3 * SEPTEMBER 2000
609
610
KIMURA
AG (a serum pepsinogen I level <70 ng/ml and a pepsinogen I-topepsinogen I1 ratio C3.0). The prevalence of AG was 11%in subjects aged 16 to 29 years, 28% in subjects aged 30 to 49 years, and 43% in subjects aged 50 to 64 years. Although the prevalences of H. pylori infection were consistently around 85% among the subjects with AG in all age subgroups, the prevalences increased with age among subjects without AG (from 26% in those aged 16 to 29 years to 61% in those aged 50 to 64 years). The author routinely examines the extent of AG endoscopically using the atrophic pattern system proposed by Kimura and Takernoto,'* which is widely used in Japan. Atrophic patterns originally were classified into 6 types according to the location of the endoscopic atrophic border in the stomach (Fig. 1). Later, type C-0 and 0-p were conceptually added, in which an endoscopic atrophic border is not yet or no longer invisible. The atrophic border is the boundary between the nonatrophic mucosa and atrophic mucosa, coincidentally being also the border between the fundic gland area and pyloric gland area, which corresponds histologically to the intermediate or transitional zone from the fundic gland to the pyloric gland territories. The atrophic border is recognized endoscopically as a borderline or transitional zone, which is observed by discriminating differences in color and height of the two adjacent areas of gastric mucosa. In the closed type C-0, no atrophic changes are found. In the C-1 type, atrophic changes are visible endoscopically localized only in the antrum. In the C-2 and C-3 types, the atrophic borders are observed running across the greater curvature approximately at the entrance of the antrum, going proximally on the lesser curvature of the lower and upper portions of the body. In the open type 0 - 2 , the atrophic border is located between the lesser curvature and the anterior and posterior wall of the body, and in the 0 - 2 type, it is recognized located just amid the anterior and posterior wall of the body. In the 03 type, the endoscopically atrophic gastric mucosa spreads far more Closed Type
Open Type
Figure 1. A classification of the endoscopic atrophic pattern.
GASTRITIS AND GASTRIC CANCER
611
extensively, the endoscopic atrophic border being located between the anterior wall and the greater curvature of the body. In the 0 - p (pangatrophic) type, the gastric mucosa is replaced entirely with endoscopically atrophic mucosa throughout the entire stomach, where the endoscopic atrophic border no longer is visible. Kimurall reported that AG extended proximally with increasing age. Satoh et all5 previously examined the extent of AG of 76 nonulcer patients with and without H. pylori infection using endoscopy and histology of biopsy specimens taken from six points in each patient (described subsequently) and reported the relationship between the age and extension of AG (Fig. 2). In H. pylori-positive patients, the moderate extent of AG (O-m) appeared from patients aged 40 to 49 years, and the severe extent of AG (0-p) appeared from patients aged 50 to 59 years. In contrast, most H. pylori-negative patients had no atrophy (C-0) or antral atrophy (C-1) only, regardless of age. AG is considered to develop in H. pylori-positive patients and progresses extensively over a period of decades. HELlCOBACTER PYLORl INFECTION AND GASTRIC CANCER
The prospective studies of Forman et al,7 Parsonnet et al,14 and Nomura et all3 showed that there was an odds ratio (OR) of 2.8 to 6.0 for the development of gastric cancer in H. pylori-positive patients. The case-control studies in Japan showed that the ORs in H.pylori-positive Japanese patients', 4, lo, were not high, however, except for one high
Figure 2. Relationship between age and atrophic pattern. O-m indicates the moderate extent of atrophic gastritis extending proximally to the middle or upper body on the lesser curvature, including endoscopic atrophic pattern of types C-3,0-1, 0-2, and 0-3. Solid circles = H. pylorkpositive; circles = H. pylori-negative.
612
KIMURA
OR of 13.3 calculated in only a young population by Kikuchi et all0 (Table 1). Watanabe et all8conducted a multivariate analysis and showed that H. pylori infection was not associated with gastric cancer (OR, 1.84; 95% confidence interval [CI], 0.59 to 5.72; P<.05), but AG was associated with a significantly increased risk for cancer (OR, 3.38; 95% CI, 1.54 to 7.42; P = .0025). Fukao et a17studied the prevalences of H. pylori infection and AG in four prefectures in Japan in relation to various standardized mortality ratios for gastric cancer, reporting that the prevalence of AG paralleled the standardized mortality ratios for gastric cancer, but the prevalences of H. pylori infection were not different among the four prefectures. These findings suggest that AG per se is associated strongly with the development of gastric cancer in Japanese patients, but H. pylori infection is not. IMPORTANCE OF BIOPSY SITES FOR THE DIAGNOSIS OF ATROPHIC GASTRITIS
Because H. pylori infection and AG are associated strongly with gastric cancer, diagnosis of these conditions is important. An examination of H. pylori and gastritis should be made, taking into consideration the extent of AG, especially in countries with high prevalences of AG, such as Japan. For the assessment of the extent of AG, and for the diagnosis of H. pylori infection, endoscopic observation including biopsy is mandatory at present. Endoscopic biopsy from the specified sites is especially important. For effective study, the number of biopsy specimens required and their location must be specified, rather than allowing the mucosa to be surveyed arbitrarily. Biopsies should be performed systematically according to their purposes. What location and number of biopsy specimens are required to assess H. pylori infection and AG? In the stomach, the immobile landmarks are the esophagogastric junction, the incisura angularis, the pyloric ring on the lesser curvature, the boundary between the fundus and the body, and the entrance to the antrum just opposite the incisura angularis on the greater curvature. The biopsy specimen should be taken from a site that is an immobile landmark or from the midpoint between
Table 1. ODDS RATIOS FOR THE ASSOCIATION OF GASTRIC CANCER WITH HELlCOBACTER PYLORl SEROPOSITIVITY IN JAPAN Author
Fukuda8 Blase? Asaka” KikUChi’O
Watanabe’* CI
= Confidence interval.
Odds Ratio (95% CI) 1.7 2.1 2.6 13.3 1.84
(1.01-2.81) (0.724.4) (1.5-4.2) (5.S35.6) (0.59-5.72)
GASTRITIS AND GASTRIC CANCER
613
the two adjacent immobile landmarks (Fig. 3). The biopsy sites suitable for the diagnosis of H. pylori infection and the assessment of the extent of AG, performing biopsies according to the Kimura system, were invesThis system was first proposed by Kimura as the tigated previ~usly.'~ specification for biopsy sites at routine endoscopic gastric biopsy (Yokohama Recommendation at the 10th Asian Pacific Congress of Gastroenterology at Yokohama, 1996). Endoscopy was performed in 76 H. pylori-positive patients with histologically confirmed chronic gastritis. The extent of AG was assessed endoscopically according to the atrophic pattern (see Fig. ,).I2 Biopsy specimens were taken from the following six specified sites: site 1, the midantrum (at a position midway between the pyloric ring and the incisura angularis); site 2, the angulus; site 3, the middle body (at a position midway between the incisura angularis and the esophagogastric junction) on the lesser curvature; site 4, the midantrum (at a position midway between the pyloric ring and the entrance to the antrum); site 5, the entrance to the antrum just opposite to the incisura angularis; and site 6, the middle body (at a position midway between the entrance to the antrum and the boundary between the corpus and the fundus) on the greater curvature (see Fig. 3). The histologic severity of H. pylori density, inflammation, activity, atrophy, and intestinal metaplasia was graded as none, mild, moderate, and marked, according to the Updated Sydney S y ~ t e mand , ~ was scored on a 0-to-3 scale. The extent of AG assessed endoscopically was modified based on the histologic findings of atrophy at the six biopsy sites (see Fig. 3). Patients were classified into the five groups according to the extent of AG: nonatrophic gastritis (NAG), antral, mild (extending beyond the incisura angularis), moderate (extending to the middle or upper body on the lesser curvature), and severe extent (spreading throughout the entire stomach) of AG. As to mucosal gland type, site 1 and site 4 revealed almost exclusively pyloric; site 3 and site 5, transitional; and site 6, fundic types in most cases (Fig. 4).H. pylori density was even in the stomach of patients
n
Incisura
Site
I .
Figure 3. Specification of biopsy site. Site 1 = midantrum; site 2 = angulus; site 3 = middle body on the lesser curvature, respectively; site 4 = midantrum; site 5 = entrance into the antrum; site 6 = middle body on the greater curvature, respectively.
614
KIMURA
Figure 4. Mucosal gland type at each biopsy site. The percentages of three mucosal gland types at each biopsy site were shown according to the extent of atrophic gastritis. NAG = nonatrophic gastritis; Antral = antral atrophic gastritis; Mild = mild extent; Mod = moderate extent; Sev = severe extent of atrophic gastritis; open bar = antral; hatched bar = transitional; shaded bar = oxyntic.
of the NAG and antral and mild AG groups but significantly higher at site 6 than that at site 1 and site 2 in the moderate and severe AG groups (Fig. 5). Inflammation score at site 6 was lower in the NAG group but significantlyhigher in the severe AG group than that at site 1 and site 2 (Fig. 6). Activity score (neutrophil infiltration) was significantly higher at site 6 than at site 1, site 2, and site 3 in the moderate and severe AG groups (Fig. 7). Atrophy was found in 67 patients (Fig. 8). In 4 of the 5 patients with antral AG, only one of the six biopsy specimens showed atrophy: 3 patients showed atrophy at site 1 and 1 at site 4. Intestinal metaplasia was found in 58 patients. The prevalences of intestinal metaplasia in the stomachs of these patients were similar at the lesser curvature (71% at site 1, 74% at site 2, and 79% at site 3) but low at the greater curvature (55% at site 4, 41% at site 5, and 7% at site 6). In 9 patients, only one of the six biopsy specimens showed intestinal metaplasia: 2 patients showed IM at site 2, 6 at site 3, and 1 at site 4. In patients with extensive intestinal metaplasia, the scores of intestinal metaplasia at site 1 and site 3 were as high as that at site 2 (mean scores, 1.5, 2, and 1.6 in the moderate AG group and 2.1, 2.3, and 2 in the severe AG group). Table 2 shows the biopsy sites suitable for the assessment of each item in the stomach. In the antrum, site 4 is suitable for H. pylori, inflammation, and activity, and site 1 is suitable for atrophy and intestinal metaplasia. In the transitional zone, site 3 is better than site 5. In the body, site 6 is suitable for all items. In the whole stomach, site 6 was most reliable for the diagnosis of H. pylori infection, and site 4 was suitable for examining the status of H. pylori colonization in the antrum.
GASTRITIS AND GASTRIC CANCER
615
Figure 5. Score of H. pyloridensity at each biopsy site. n = number of patients; **P
616
KIMURA
Figure 6. Score of inflammation at each biopsy site. n = number of patients; *P<0.05. One biopsy at site 5 in the antral and mild extent of atrophic gastritis groups, and one at site 6 in the moderate extent of atrophic gastritis group were not large enough to permit accurate grading.
GASTRITIS AND GASTRIC CANCER
617
Figure 7. Score of activity at each biopsy site. n = number of patients; *P<0.05; **P
618
KIMURA
Figure 8. Score of atrophy at each biopsy site. n = number of patients; *P<0.05;**P
GASTRITIS AND GASTRIC CANCER
619
Table 2. OPTIMAL BIOPSY SITES Optimal Site for Diagnosis of
H.pylori Inflammation Activity Atrophy Intestinal metaplasia
Whole Stomach 6 6* or It 6 1 1
Antrum
4 land4 4
1 1
Transitional Zone
Body
3and5 3 and 5 3 and 5 3 3
6 6 6 6 6
*For patients with the severe extent of atrophic gastritis. tFor patients with nonatrophic gastritis.
Site 1, site 3, and site 6 were suitable for the assessment of the extent of AG. Biopsy specimens for an accurate assessment of H. pylori infection and AG should be taken from the following four sites: the midantrum (site 1)and middle body (site 3) on the lesser curvature and the midantrum (site 4) and middle body (site 6) on the greater curvature. The biopsy sites recommended for the optimal diagnosis of H. pylori are site 4 and site 6 on the greater curvature, and for assessment of the extent of AG, one also should sample site 1 and site 3 on the lesser curvature and site 6 on the greater curvature. SUMMARY
In Japan, high prevalences of H. pylori infection and AG are found, which seem to be causes for the high incidence of gastric cancer. H. pylori infection is not a sole cause of gastric cancer, however. Environmental factors, including a diet high in salt, also might be involved. For accurate diagnosis of H. pylori infection and AG, it is important to determine the locations from which the biopsy specimens are taken. Comparison of the severity of AG between different countries should be made using the same biopsy sampling system and grading criteria. Bayerdoerffer et a13showed antral predominance of H. pylori colonization, but Genta et a19 reported more even distribution in the stomach. The authors’ patients with NAG and those with antral and mild AG had similar scores of density at the six biopsy sites, whereas those with more extensive AG had the highest score at the middle body on the greater curvature. If only one biopsy specimen should be taken for the diagnosis of H. pylori infection in patients with extensive AG, the middle body of the greater curvature (site 6) would be the most suitable site. Western researchers reported that the grade of chronic3, and acute inflammation9was significantly lower in the body than in the antrum. The authors’ findings of patients with NAG were similar to those of Western researchers, but those with atrophy were not. Grades of inflammation and activity in the antrum decrease with the extension of AG.16
620
KIMURA
In the authors’ patients, the earliest finding of the development of atrophy was found at site 1, and this was the most predominant site of atrophy in patients with extensive AG. Extension of AG into the body was determined by the finding of a biopsy specimen taken at site 3. The earliest finding of the development of intestinal metaplasia was found mostly at the middle body of the lesser curvature and next at the incisura angularis. In patients with extensive intestinal metaplasia, the scores of intestinal metaplasia at site 1 and site 3 were as high as the score at site 2. Site 6 was the last place for the invasion from the extension of AG. If biopsy specimens are taken from site 1, site 3, and site 6, they are enough for the assessment of the extent of AG and intestinal metaplasia. The Updated Sydney System5 recommended five biopsy sites including the incisura angularis, and its usefulness was coincidentally supported in Japan with a high prevalence of AG seen in results.
References 1. Asaka M, Kimura T, Kato M, et al: Possible role of Helicobacter pylori infection in early gastric cancer development. Cancer 73:2691-2694, 1994 2. Asaka M, Kimura T, Kudo M, et al: Relationship of Helicobacter pylori to serum pepsinogens in an asymptomatic Japanese population. Gastroenterology 102:760-766, 1992 3. Bayerdoerffer E, Lehn N, Hatz R, et al: Difference in expression of Helicobacter pylori gastritis in antrum and body. Gastroenterology 102:1575-1582, 1992 4. Blaser MJ, Kobayashi K, Cover TL, et al: Helicobacter pylori infection in Japanese patients with adenocarcinoma of the stomach. Int J Cancer 55:799-802, 1993 5. Dixon MF, Genta RM, Yardley JH, et al: Classification and grading of gastritis: The updated Sydney system. Am J Surg Pathol 20:1161-1181, 1996 6. Forman D, Newel1 DG, Fullerton F, et al: Association between infection with Helicobacter pylori and risk of gastric cancer: Evidence from a prospective investigation. BMJ 3021302-1305, 1991 7. Fukao A, Komatsu S, Tsubono Y, et al: Helicobacter pylori infection and atrophic gastritis among Japanese blood donors: A cross-sectional study. Cancer Causes Control 4:307-312, 1993 8. Fukuda H, Saito D, Hayashi S, et al: Helicobacter pylori infection, serum pepsinogen level and gastric cancer: A case-control study in Japan. Jpn J Cancer Res 8636471, 1995 9. Genta RM, Huberman RM, Graham DY The gastric cardia in Helicobacter pylori infection. Hum Pathol25:915-919, 1994 10. Kikuchi S, Wada 0, Nakajima T, et al: Serum anti-Helicobacter pylori antibody and gastric carcinoma among young adults. Cancer 752789-2793, 1995 11. Kimura K: Chronological transition of the fundic-pyloric border determined by stepwise biopsy of the lesser and greater curvatures of the stomach. Gastroenterology 63:584-592, 1972 12. Kimura K, Takemoto T An endoscopic recognition of the atrophic border and its significance in chronic gastritis. Endoscopy 3:87-97, 1969 13. Nomura A, Stemmermann GN, Chyou PH, et al: Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii. N Engl J Med 325:1132-1136, 1991 14. Parsonnet J, Friedman GD, Vandersteen DP, et al: Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 325:1127-1131, 1991 15. Satoh K, Kihira K, Kimura K: Extension of atrophic gastritis in patients with and without Helicobacter pylori infection. Stomach Intestine 33:1131-1136, 1998 16. Satoh K, Kimura K, Taniguchi Y, et al: Distribution of inflammation and atrophy in
GASTRITIS AND GASTRIC CANCER
621
the stomach of Helicobacter pylori-positive and -negative patients with chronic gastritis. Am J Gastroenterol 91:963-969, 1996 17. Satoh K, Kimura K, Taniguchi Y, et al: Biopsy sites suitable for the diagnosis of Helicobacter pylori infection and the assessment of the extent of atrophic gastritis. Am J Gastroenterol 93:569-573, 1998 18. Watanabe Y, Kurata JH, Mizuno S, et al: Helicobacter pylori infection and gastric cancer: A nested case-control study in a rural area of Japan. Dig Dis Sci 421383-1387, 1997 Address reprint requests to Ken Kimura, MD Department of Gastroenterology Jichi Medical School Yakushiji, Tochigi 3294498 Japan
0889-8553 / 00 $15.00
+ .OO
GASTRITIS AND GASTRIC CANCER Asia Ken Kimura, MD
Japan has the highest annual incidence of gastric cancer in the world. The prevalences of Helicobacter pylori infection and atrophic gastritis (AG) in Japan are also high. H. pylori infection causes chronic gastritis, and it leads to development and progression of AG and intestinal metaplasia, which are considered precursor lesions of gastric cancer. This article describes the relationship of H. pylori infection and AG and gastric cancer in Japan. PREVALENCE OF HELICOBACTER PYLORI INFECTION AND ATROPHIC GASTRITIS
In 1992, Asaka et a12 reported the prevalence of H. pylori infection in the asymptomatic Japanese population using serum anti-H. pylori IgG antibody. The prevalence was less than 10% in persons aged 0 to 9 years, was about 20% in persons aged 10 to 19 years, increased with age, and reached the plateau of 80% in persons aged 40 to 49 years. The prevalence in persons born after 1950 increased at approximately 1%per year, and the prevalence in persons born before 1949 was high, relatively constant, and approximately double of that of the 30- to 40-year age group. There are two cohorts that differ in the risk for H. pylori in Japan. In 1993, Fukao et a17 studied serum samples of 1815 healthy blood donors and reported the prevalence of H. pylori infection and presumed
From the Department of Gastroenterology, Jichi Medical School, Tochigi, Japan
GASTROENTEROLOGY CLINICS OF NORTH AMERICA
-
VOLUME 29 NUMBER 3 * SEPTEMBER 2000
609
610
KIMURA
AG (a serum pepsinogen I level <70 ng/ml and a pepsinogen I-topepsinogen I1 ratio C3.0). The prevalence of AG was 11%in subjects aged 16 to 29 years, 28% in subjects aged 30 to 49 years, and 43% in subjects aged 50 to 64 years. Although the prevalences of H. pylori infection were consistently around 85% among the subjects with AG in all age subgroups, the prevalences increased with age among subjects without AG (from 26% in those aged 16 to 29 years to 61% in those aged 50 to 64 years). The author routinely examines the extent of AG endoscopically using the atrophic pattern system proposed by Kimura and Takernoto,'* which is widely used in Japan. Atrophic patterns originally were classified into 6 types according to the location of the endoscopic atrophic border in the stomach (Fig. 1). Later, type C-0 and 0-p were conceptually added, in which an endoscopic atrophic border is not yet or no longer invisible. The atrophic border is the boundary between the nonatrophic mucosa and atrophic mucosa, coincidentally being also the border between the fundic gland area and pyloric gland area, which corresponds histologically to the intermediate or transitional zone from the fundic gland to the pyloric gland territories. The atrophic border is recognized endoscopically as a borderline or transitional zone, which is observed by discriminating differences in color and height of the two adjacent areas of gastric mucosa. In the closed type C-0, no atrophic changes are found. In the C-1 type, atrophic changes are visible endoscopically localized only in the antrum. In the C-2 and C-3 types, the atrophic borders are observed running across the greater curvature approximately at the entrance of the antrum, going proximally on the lesser curvature of the lower and upper portions of the body. In the open type 0 - 2 , the atrophic border is located between the lesser curvature and the anterior and posterior wall of the body, and in the 0 - 2 type, it is recognized located just amid the anterior and posterior wall of the body. In the 03 type, the endoscopically atrophic gastric mucosa spreads far more Closed Type
Open Type
Figure 1. A classification of the endoscopic atrophic pattern.
GASTRITIS AND GASTRIC CANCER
611
extensively, the endoscopic atrophic border being located between the anterior wall and the greater curvature of the body. In the 0 - p (pangatrophic) type, the gastric mucosa is replaced entirely with endoscopically atrophic mucosa throughout the entire stomach, where the endoscopic atrophic border no longer is visible. Kimurall reported that AG extended proximally with increasing age. Satoh et all5 previously examined the extent of AG of 76 nonulcer patients with and without H. pylori infection using endoscopy and histology of biopsy specimens taken from six points in each patient (described subsequently) and reported the relationship between the age and extension of AG (Fig. 2). In H. pylori-positive patients, the moderate extent of AG (O-m) appeared from patients aged 40 to 49 years, and the severe extent of AG (0-p) appeared from patients aged 50 to 59 years. In contrast, most H. pylori-negative patients had no atrophy (C-0) or antral atrophy (C-1) only, regardless of age. AG is considered to develop in H. pylori-positive patients and progresses extensively over a period of decades. HELlCOBACTER PYLORl INFECTION AND GASTRIC CANCER
The prospective studies of Forman et al,7 Parsonnet et al,14 and Nomura et all3 showed that there was an odds ratio (OR) of 2.8 to 6.0 for the development of gastric cancer in H. pylori-positive patients. The case-control studies in Japan showed that the ORs in H.pylori-positive Japanese patients', 4, lo, were not high, however, except for one high
Figure 2. Relationship between age and atrophic pattern. O-m indicates the moderate extent of atrophic gastritis extending proximally to the middle or upper body on the lesser curvature, including endoscopic atrophic pattern of types C-3,0-1, 0-2, and 0-3. Solid circles = H. pylorkpositive; circles = H. pylori-negative.
612
KIMURA
OR of 13.3 calculated in only a young population by Kikuchi et all0 (Table 1). Watanabe et all8conducted a multivariate analysis and showed that H. pylori infection was not associated with gastric cancer (OR, 1.84; 95% confidence interval [CI], 0.59 to 5.72; P<.05), but AG was associated with a significantly increased risk for cancer (OR, 3.38; 95% CI, 1.54 to 7.42; P = .0025). Fukao et a17studied the prevalences of H. pylori infection and AG in four prefectures in Japan in relation to various standardized mortality ratios for gastric cancer, reporting that the prevalence of AG paralleled the standardized mortality ratios for gastric cancer, but the prevalences of H. pylori infection were not different among the four prefectures. These findings suggest that AG per se is associated strongly with the development of gastric cancer in Japanese patients, but H. pylori infection is not. IMPORTANCE OF BIOPSY SITES FOR THE DIAGNOSIS OF ATROPHIC GASTRITIS
Because H. pylori infection and AG are associated strongly with gastric cancer, diagnosis of these conditions is important. An examination of H. pylori and gastritis should be made, taking into consideration the extent of AG, especially in countries with high prevalences of AG, such as Japan. For the assessment of the extent of AG, and for the diagnosis of H. pylori infection, endoscopic observation including biopsy is mandatory at present. Endoscopic biopsy from the specified sites is especially important. For effective study, the number of biopsy specimens required and their location must be specified, rather than allowing the mucosa to be surveyed arbitrarily. Biopsies should be performed systematically according to their purposes. What location and number of biopsy specimens are required to assess H. pylori infection and AG? In the stomach, the immobile landmarks are the esophagogastric junction, the incisura angularis, the pyloric ring on the lesser curvature, the boundary between the fundus and the body, and the entrance to the antrum just opposite the incisura angularis on the greater curvature. The biopsy specimen should be taken from a site that is an immobile landmark or from the midpoint between
Table 1. ODDS RATIOS FOR THE ASSOCIATION OF GASTRIC CANCER WITH HELlCOBACTER PYLORl SEROPOSITIVITY IN JAPAN Author
Fukuda8 Blase? Asaka” KikUChi’O
Watanabe’* CI
= Confidence interval.
Odds Ratio (95% CI) 1.7 2.1 2.6 13.3 1.84
(1.01-2.81) (0.724.4) (1.5-4.2) (5.S35.6) (0.59-5.72)
GASTRITIS AND GASTRIC CANCER
613
the two adjacent immobile landmarks (Fig. 3). The biopsy sites suitable for the diagnosis of H. pylori infection and the assessment of the extent of AG, performing biopsies according to the Kimura system, were invesThis system was first proposed by Kimura as the tigated previ~usly.'~ specification for biopsy sites at routine endoscopic gastric biopsy (Yokohama Recommendation at the 10th Asian Pacific Congress of Gastroenterology at Yokohama, 1996). Endoscopy was performed in 76 H. pylori-positive patients with histologically confirmed chronic gastritis. The extent of AG was assessed endoscopically according to the atrophic pattern (see Fig. ,).I2 Biopsy specimens were taken from the following six specified sites: site 1, the midantrum (at a position midway between the pyloric ring and the incisura angularis); site 2, the angulus; site 3, the middle body (at a position midway between the incisura angularis and the esophagogastric junction) on the lesser curvature; site 4, the midantrum (at a position midway between the pyloric ring and the entrance to the antrum); site 5, the entrance to the antrum just opposite to the incisura angularis; and site 6, the middle body (at a position midway between the entrance to the antrum and the boundary between the corpus and the fundus) on the greater curvature (see Fig. 3). The histologic severity of H. pylori density, inflammation, activity, atrophy, and intestinal metaplasia was graded as none, mild, moderate, and marked, according to the Updated Sydney S y ~ t e mand , ~ was scored on a 0-to-3 scale. The extent of AG assessed endoscopically was modified based on the histologic findings of atrophy at the six biopsy sites (see Fig. 3). Patients were classified into the five groups according to the extent of AG: nonatrophic gastritis (NAG), antral, mild (extending beyond the incisura angularis), moderate (extending to the middle or upper body on the lesser curvature), and severe extent (spreading throughout the entire stomach) of AG. As to mucosal gland type, site 1 and site 4 revealed almost exclusively pyloric; site 3 and site 5, transitional; and site 6, fundic types in most cases (Fig. 4).H. pylori density was even in the stomach of patients
n
Incisura
Site
I .
Figure 3. Specification of biopsy site. Site 1 = midantrum; site 2 = angulus; site 3 = middle body on the lesser curvature, respectively; site 4 = midantrum; site 5 = entrance into the antrum; site 6 = middle body on the greater curvature, respectively.
614
KIMURA
Figure 4. Mucosal gland type at each biopsy site. The percentages of three mucosal gland types at each biopsy site were shown according to the extent of atrophic gastritis. NAG = nonatrophic gastritis; Antral = antral atrophic gastritis; Mild = mild extent; Mod = moderate extent; Sev = severe extent of atrophic gastritis; open bar = antral; hatched bar = transitional; shaded bar = oxyntic.
of the NAG and antral and mild AG groups but significantly higher at site 6 than that at site 1 and site 2 in the moderate and severe AG groups (Fig. 5). Inflammation score at site 6 was lower in the NAG group but significantlyhigher in the severe AG group than that at site 1 and site 2 (Fig. 6). Activity score (neutrophil infiltration) was significantly higher at site 6 than at site 1, site 2, and site 3 in the moderate and severe AG groups (Fig. 7). Atrophy was found in 67 patients (Fig. 8). In 4 of the 5 patients with antral AG, only one of the six biopsy specimens showed atrophy: 3 patients showed atrophy at site 1 and 1 at site 4. Intestinal metaplasia was found in 58 patients. The prevalences of intestinal metaplasia in the stomachs of these patients were similar at the lesser curvature (71% at site 1, 74% at site 2, and 79% at site 3) but low at the greater curvature (55% at site 4, 41% at site 5, and 7% at site 6). In 9 patients, only one of the six biopsy specimens showed intestinal metaplasia: 2 patients showed IM at site 2, 6 at site 3, and 1 at site 4. In patients with extensive intestinal metaplasia, the scores of intestinal metaplasia at site 1 and site 3 were as high as that at site 2 (mean scores, 1.5, 2, and 1.6 in the moderate AG group and 2.1, 2.3, and 2 in the severe AG group). Table 2 shows the biopsy sites suitable for the assessment of each item in the stomach. In the antrum, site 4 is suitable for H. pylori, inflammation, and activity, and site 1 is suitable for atrophy and intestinal metaplasia. In the transitional zone, site 3 is better than site 5. In the body, site 6 is suitable for all items. In the whole stomach, site 6 was most reliable for the diagnosis of H. pylori infection, and site 4 was suitable for examining the status of H. pylori colonization in the antrum.
GASTRITIS AND GASTRIC CANCER
615
Figure 5. Score of H. pyloridensity at each biopsy site. n = number of patients; **P
616
KIMURA
Figure 6. Score of inflammation at each biopsy site. n = number of patients; *P<0.05. One biopsy at site 5 in the antral and mild extent of atrophic gastritis groups, and one at site 6 in the moderate extent of atrophic gastritis group were not large enough to permit accurate grading.
GASTRITIS AND GASTRIC CANCER
617
Figure 7. Score of activity at each biopsy site. n = number of patients; *P<0.05; **P
618
KIMURA
Figure 8. Score of atrophy at each biopsy site. n = number of patients; *P<0.05;**P
GASTRITIS AND GASTRIC CANCER
619
Table 2. OPTIMAL BIOPSY SITES Optimal Site for Diagnosis of
H.pylori Inflammation Activity Atrophy Intestinal metaplasia
Whole Stomach 6 6* or It 6 1 1
Antrum
4 land4 4
1 1
Transitional Zone
Body
3and5 3 and 5 3 and 5 3 3
6 6 6 6 6
*For patients with the severe extent of atrophic gastritis. tFor patients with nonatrophic gastritis.
Site 1, site 3, and site 6 were suitable for the assessment of the extent of AG. Biopsy specimens for an accurate assessment of H. pylori infection and AG should be taken from the following four sites: the midantrum (site 1)and middle body (site 3) on the lesser curvature and the midantrum (site 4) and middle body (site 6) on the greater curvature. The biopsy sites recommended for the optimal diagnosis of H. pylori are site 4 and site 6 on the greater curvature, and for assessment of the extent of AG, one also should sample site 1 and site 3 on the lesser curvature and site 6 on the greater curvature. SUMMARY
In Japan, high prevalences of H. pylori infection and AG are found, which seem to be causes for the high incidence of gastric cancer. H. pylori infection is not a sole cause of gastric cancer, however. Environmental factors, including a diet high in salt, also might be involved. For accurate diagnosis of H. pylori infection and AG, it is important to determine the locations from which the biopsy specimens are taken. Comparison of the severity of AG between different countries should be made using the same biopsy sampling system and grading criteria. Bayerdoerffer et a13showed antral predominance of H. pylori colonization, but Genta et a19 reported more even distribution in the stomach. The authors’ patients with NAG and those with antral and mild AG had similar scores of density at the six biopsy sites, whereas those with more extensive AG had the highest score at the middle body on the greater curvature. If only one biopsy specimen should be taken for the diagnosis of H. pylori infection in patients with extensive AG, the middle body of the greater curvature (site 6) would be the most suitable site. Western researchers reported that the grade of chronic3, and acute inflammation9was significantly lower in the body than in the antrum. The authors’ findings of patients with NAG were similar to those of Western researchers, but those with atrophy were not. Grades of inflammation and activity in the antrum decrease with the extension of AG.16
620
KIMURA
In the authors’ patients, the earliest finding of the development of atrophy was found at site 1, and this was the most predominant site of atrophy in patients with extensive AG. Extension of AG into the body was determined by the finding of a biopsy specimen taken at site 3. The earliest finding of the development of intestinal metaplasia was found mostly at the middle body of the lesser curvature and next at the incisura angularis. In patients with extensive intestinal metaplasia, the scores of intestinal metaplasia at site 1 and site 3 were as high as the score at site 2. Site 6 was the last place for the invasion from the extension of AG. If biopsy specimens are taken from site 1, site 3, and site 6, they are enough for the assessment of the extent of AG and intestinal metaplasia. The Updated Sydney System5 recommended five biopsy sites including the incisura angularis, and its usefulness was coincidentally supported in Japan with a high prevalence of AG seen in results.
References 1. Asaka M, Kimura T, Kato M, et al: Possible role of Helicobacter pylori infection in early gastric cancer development. Cancer 73:2691-2694, 1994 2. Asaka M, Kimura T, Kudo M, et al: Relationship of Helicobacter pylori to serum pepsinogens in an asymptomatic Japanese population. Gastroenterology 102:760-766, 1992 3. Bayerdoerffer E, Lehn N, Hatz R, et al: Difference in expression of Helicobacter pylori gastritis in antrum and body. Gastroenterology 102:1575-1582, 1992 4. Blaser MJ, Kobayashi K, Cover TL, et al: Helicobacter pylori infection in Japanese patients with adenocarcinoma of the stomach. Int J Cancer 55:799-802, 1993 5. Dixon MF, Genta RM, Yardley JH, et al: Classification and grading of gastritis: The updated Sydney system. Am J Surg Pathol 20:1161-1181, 1996 6. Forman D, Newel1 DG, Fullerton F, et al: Association between infection with Helicobacter pylori and risk of gastric cancer: Evidence from a prospective investigation. BMJ 3021302-1305, 1991 7. Fukao A, Komatsu S, Tsubono Y, et al: Helicobacter pylori infection and atrophic gastritis among Japanese blood donors: A cross-sectional study. Cancer Causes Control 4:307-312, 1993 8. Fukuda H, Saito D, Hayashi S, et al: Helicobacter pylori infection, serum pepsinogen level and gastric cancer: A case-control study in Japan. Jpn J Cancer Res 8636471, 1995 9. Genta RM, Huberman RM, Graham DY The gastric cardia in Helicobacter pylori infection. Hum Pathol25:915-919, 1994 10. Kikuchi S, Wada 0, Nakajima T, et al: Serum anti-Helicobacter pylori antibody and gastric carcinoma among young adults. Cancer 752789-2793, 1995 11. Kimura K: Chronological transition of the fundic-pyloric border determined by stepwise biopsy of the lesser and greater curvatures of the stomach. Gastroenterology 63:584-592, 1972 12. Kimura K, Takemoto T An endoscopic recognition of the atrophic border and its significance in chronic gastritis. Endoscopy 3:87-97, 1969 13. Nomura A, Stemmermann GN, Chyou PH, et al: Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii. N Engl J Med 325:1132-1136, 1991 14. Parsonnet J, Friedman GD, Vandersteen DP, et al: Helicobacter pylori infection and the risk of gastric carcinoma. N Engl J Med 325:1127-1131, 1991 15. Satoh K, Kihira K, Kimura K: Extension of atrophic gastritis in patients with and without Helicobacter pylori infection. Stomach Intestine 33:1131-1136, 1998 16. Satoh K, Kimura K, Taniguchi Y, et al: Distribution of inflammation and atrophy in
GASTRITIS AND GASTRIC CANCER
621
the stomach of Helicobacter pylori-positive and -negative patients with chronic gastritis. Am J Gastroenterol 91:963-969, 1996 17. Satoh K, Kimura K, Taniguchi Y, et al: Biopsy sites suitable for the diagnosis of Helicobacter pylori infection and the assessment of the extent of atrophic gastritis. Am J Gastroenterol 93:569-573, 1998 18. Watanabe Y, Kurata JH, Mizuno S, et al: Helicobacter pylori infection and gastric cancer: A nested case-control study in a rural area of Japan. Dig Dis Sci 421383-1387, 1997 Address reprint requests to Ken Kimura, MD Department of Gastroenterology Jichi Medical School Yakushiji, Tochigi 3294498 Japan