Serological and direct diagnosis of Helicobacter pylori in gastric carcinoma: a case-control study

This study evaluated the sensitivity of serological and direct methods for the diagnosis of Helicobacter pylori infection in 127 patients with gastric carcinoma and in 127 controls without this disease, matched for age and sex. Antral and oxyntic mucosal specimens were obtained from all patients, at operation in patients with gastric carcinoma and at endoscopy from controls. The urease test, microscopy of stained smears and culture for H. pylori were performed on all specimens. Sera from all patients were tested for antibodies to H. pylori by a highly sensitive and specific IgG-ELISA. Culture, urease test, stained smear and ELISA were significantly less sensitive in the patients with gastric carcinoma than in control subjects. However, the combination of several methods improved the diagnosis of H. pylori infection in the gastric carcinoma group. Infection was significantly more frequent in the gastric carcinoma patients than in the controls. H. pylori infection was associated with an increased risk of developing gastric carcinoma.     

Helicobacter pylori in tumor tissues of patients with advanced gastric adenocarcinoma: high prevalence but failure to detect integration

Helicobacter pylori has been known to be associated with gastric adenocarcinoma by case control studies. However, significant portion of patients with gastric carcinoma are negative for H. pylori by serological test. To further detect the presence of H. pylori infection in serum and tissue of patients with gastric adenocarcinoma, paired tissues and serum samples from 32 patients with gastric adenocarcinoma were tested. Antibodies to H. pylori were tested by an enzyme linked immunosorbent assay (ELISA) and a Western blot analysis. H. pylori in tumor and non-tumor parts of gastric tissues were examined by histology and polymerase chain reaction (PCR). For serum antibody, eighteen (56%) of these patients were positive by ELISA while 24 (75%) were positive by Western blot. For tissue H. pylori genome, 14 were positive by histology while 28 (87%) were positive by PCR. Southern blot analysis of both tumor and non-tumor tissues revealed no evidence of integration of H. pylori DNA in the human genomes. These results suggest that H. pylori infection can be detected in most patients with gastric adenocarcinoma, and PCR and Western blot can further identify seronegative patients.     

Evaluation of enzyme immunoassay for detection of salivary antibody to Helicobacter pylori.

The Helisal test is a quantitative enzyme immunoassay for the measurement of Helicobacter pylori-specific immunoglobulin G antibodies in saliva. This test was evaluated in comparison with culture and histopathologic examination of gastric biopsy specimens obtained from 195 patients who underwent 200 endoscopic procedures for the investigation of gastrointestinal symptoms. Forty-one (21%) patients were found to have peptic ulcer disease, and one other patient had a gastric carcinoma. H. pylori was detected in gastric biopsy specimens obtained from 98 (49%) of the procedures. The sensitivity, specificity, and positive and negative predictive values of the Helisal test were 81, 75, 76, and 80%, respectively. The test was negative for 16 (38%) of the 42 patients with peptic ulcer disease or a gastric malignancy diagnosed at endoscopy. These results suggest that the Helisal assay is only moderately accurate for the detection of H. pylori infection in symptomatic patients.     

Correlation of serum antibody titres with invasive methods for rapid detection of Helicobacter pylori infections in symptomatic children

Helicobacter pylori (H. pylori) is causally associated with peptic ulcer disease and gastric carcinoma. Typically, children get infected during the first decade of life, but diseases associated with H. pylori are seen mainly in adults. Multiple diagnostic methods are available for the detection of H. pylori infection. The aim of this study was to evaluate the correlation and diagnostic accuracy of three invasive methods [rapid urease test (RUT), histology and bacterial culture] and one non-invasive method (IgG serology) for diagnosis of H. pylori infection in a prospective cohort study conducted on 50 symptomatic children between two and eighteen years of age. Endoscopies with gastric biopsies were performed for RUT, culture and histopathological examination, respectively. IgG antibodies were measured in patient sera using a commercially available enzyme-linked immunosorbent assay (ELISA). RUT and positive H. pylori IgG antibodies were concordant in 88% (44/50) of patients. Both tests were negative in 32% (16/50), and both were positive in 56% (28/50). Disagreement occurred in 12% (6/50) of the patients: three of them (6%) had positive RUT and negative H. pylori IgG, and another three (6%) had negative RUT and positive H. pylori IgG. A combination of RUT with non-invasive serology constituted the optimum approach to the diagnosis of H. pylori infection in symptomatic children. The non-invasive serological test (ELISA) could not be used alone as the gold standard because it cannot distinguish between active and recently treated infection; and bacterial culture could not be used alone because of its low sensitivity.     

Helicobacter pylori infection and chronic gastritis in gastric cancer.

AIMS: To investigate the prevalence of Helicobacter pylori associated chronic gastritis in patients with gastric cancer. METHODS: Serum IgG antibodies for H pylori were determined in 54 consecutive patients with gastric carcinoma. The prevalence of H pylori in gastric mucosa was also examined histologically (modified Giemsa) in 32 patients from whom adequate biopsy specimens of the antrum and corpus were available. Thirty five patients with gastrointestinal tumours outside the stomach and 48 with non-gastrointestinal malignancies served as controls. RESULTS: Of the 54 patients, 38 (70%) had H pylori antibodies (IgG) in their serum (three additional patients had H pylori antibodies IgA, class specific but not IgG specific). This prevalence was significantly higher (p less than 0.05) than that (49%) in the 35 controls. No differences in prevalence of H pylori antibodies were found between gastric cancer cases of intestinal (IGCA) or diffuse (DGCA) type, both these types showing H pylori antibodies (IgG) in 71% of the patients. In the subgroup of 32 subjects, five patients had normal gastric mucosa and four showed corpus limited atrophy ("pernicious anaemia type" atrophy of type A). All of these nine patients had no evidence of current or previous H pylori infection in serum (no IgG antibodies) or in tissue sections (negative Giemsa staining). The remaining 23 patients had antral or pangastritis, and all had evidence of current or previous H pylori infection. CONCLUSIONS: H pylori associated chronic gastritis was the associated disease in 75% of the patients with gastric cancer occurring equally often in both IGCA and DGCA groups. About 25% of cases seem to have a normal stomach or severe corpus limited atrophy, neither of which showed evidence of concomitant H pylori infection.     

Giardia lamblia trophozoites in gastric biopsies

To assess the prevalence of gastric giardiasis in gastric biopsies of patients with carcinoma stomach and in patients taking treatment for duodenal ulcer. Gastric biopsy specimens from 54 patients of carcinoma stomach and 100 antral biopsies from patients taking treatment for duodenal ulcer were included in the study. Sections were stained with haematoxylin and eosin, methylene blue and May Grunwald-Giemsa stains and examined for presence of Giardia lamblia trophozoites. Eight out of 54 (14.9%) biopsies of gastric carcinoma patients harboured trophozoites of Giardia lamblia. Associated H. pylori infection was present in all biopsies (8/8; 100%). Atrophy and intestinal metaplasia was present in 62.5% (5/8) and 25% (2/8) cases respectively. Sections from seven out of 35 patients (20%) taking treatment for duodenal ulcer showed presence of G. lamblia. H. pylori infection, gastritis and atrophy were found in 85.7% (6/7), 71.4% (5/7) and 28.6% (2/7) cases respectively. First gastric biopsy in these patients was negative for G. lamblia but 2nd and 3rd biopsies were positive. A careful search for G. lamblia trophozoites should be made while examining the gastric biopsies, especially in patients with carcinoma stomach, intestinal metaplasia, atrophic gastritis and those taking treatment for duodenal ulcer. This may help in indirect diagnosis of clinically unsuspected cases of intestinal giardiasis and may explain persistence of vague upper gastrointestinal tract (UGIT) symptoms despite clearance of H. pylori in patients on anti-ulcer therapy.     

Expression of Lewis antigens and their precursors in gastric mucosa: relationship with Helicobacter pylori infection and gastric carcinogenesis

Lewis (Le)-associated antigens are carbohydrates that are related biochemically to the ABO blood groups, and may have a role in Helicobacter pylori adherence. To evaluate their relationship to clinicopathological outcomes, gastric Le expression, including type 1 precursor, type 1 H, Le(a), Le(b), Le(x), Le(y) and sialylated Le(a) (CA19-9), was evaluated immunohistochemically in 233 gastric biopsy specimens obtained at routine gastroscopy. Expression was also investigated in gastric tissues showing chronic gastritis, intestinal metaplasia, and carcinoma from 42 patients with gastric cancer. A polymerase chain reaction was performed for H. pylori and the bacterial babA2 gene. We identified type 1 precursor expression in 34.3%, type 1 H in 55.8%, Le(a) in 44.2%, Le(b) in 82.0%, Le(x) in 44.2%, Le(y) 56.7%, and CA19-9 in 16.3% of the 233 gastric biopsy specimens. Expression of type 1 H, Le(b), and CA19-9 was significantly associated with H. pylori infection and histological features (p < 0.05), and expression of type 1 H was an independent predictive factor for H. pylori infection by multivariate logistic regression (p = 0.020). Positivity for the babA2 genotype correlated significantly with H. pylori infection and type 1 H expression in gastric biopsy specimens (p < 0.05). The babA2 genotype was more frequent in gastric mucosa from the gastric cancer patients than in gastric biopsy specimens from routine gastroscopy (p = 0.009). In the 42 gastric cancer patients, the frequency of type 1 precursor, Le(a), and Le(x) expression was significantly higher in intestinal metaplasia and carcinoma than in chronic gastritis (p < 0.05), but the frequency of type 1 H and Le(b) expression was significantly lower in intestinal metaplasia and carcinoma (p < 0.05). In conclusion, Le expression, especially that of type 1 H, was significantly associated with clinicopathological features. In gastric cancer patients, Le expression was altered in intestinal metaplasia and carcinoma in comparison with chronic gastritis.     

Helicobacter pylori infection at a university hospital in Saudi Arabia: prevalence, comparison of diagnostic modalities and endoscopic findings

The presence of Helicobacter pylori (H. pylori) was examined in 491 sequential patients, complaining mainly of epigastric pain, by three biopsy-based methods (rapid urease, histology, and culture), and by a serological test, enzyme immunosorbent assay, (ELISA). H. pylori was detected in 341 (70%) of 491 patients examined by histology, 287 (59%) by rapid urease test, whereas 385 (78%) were seropositive for H. pylori immunoglobulins by ELISA. None of the test methods used was independently sufficient to make an etiologic diagnosis of H. pylori infection. The endoscopic findings revealed that 315 (69%) of 456 patients with non-ulcer dyspepsia, 17 (74%) of 23 patients with duodenal ulcer, 7 (78%) of 9 patients with gastric ulcer, and 2 (67%) of 3 patients with gastric cancer were H. pylori positive. No statistically significant correlation was found between the endoscopic and the histopathological findings. A significant correlation was found between H. pylori infection and the histopathological gradings of gastritis (P < 0.001).     

Helicobacter pylori infection and gastric carcinoma among Japanese Americans in Hawaii

BACKGROUND. Helicobacter pylori are gram-negative spiral bacteria that are associated with chronic gastritis, a known precursor of gastric carcinoma. Persons at high risk for gastric carcinoma have been shown to have a high prevalence of H. pylori infection. METHODS. We studied the relation of H. pylori infection and gastric carcinoma in a cohort of Japanese American men living in Hawaii. The 5908 men were enrolled and examined from 1967 to 1970. By 1989, 109 cases of pathologically confirmed gastric carcinoma had been identified. The store serum of each patient with gastric carcinoma and of each matched control subject were tested for the presence of serum IgG antibody to H. pylori. RESULTS. Ninety-four percent of the men with gastric carcinoma and 76 percent of the matched control subjects had a positive test for H. pylori antibodies, for an odds ratio of 6.0 (95 percent confidence interval, 2.1 to 17.3). As the level of antibody to H. pylori increased, there was a progressive increase in the risk of gastric carcinoma (P for trend = 0.0009). The association was strong even for men in whom the diagnosis was made 10 or more years after the serum sample was obtained (odds ratio, 10.5; 95 percent confidence interval, 2.5 to 44.8). CONCLUSIONS. Infection with H. pylori is strongly associated with an increased risk of gastric carcinoma. However, most persons infected with H. pylori will never have gastric carcinoma. Therefore, other factors that increase the risk of gastric carcinoma among persons infected with H. pylori need to be identified.     

Helicobacter pylori and gastric carcinoma

A retrospective study was performed on gastric carcinomas to establish the prevalence of Helicobacter pylori infection in gastric epithelium adjacent to the tumour. A total of 105 carcinomas were studied. The overall prevalence of Helicobacter pylori infection was 59%. The prevalence in different age cohorts from patients with gastric carcinoma was compared with that in patients suffering from non-ulcer dyspepsia and, based on serological testing, with that in healthy blood donors. The presence of Helicobacter pylori in cancer patients aged 41-50 and 51-60 was significantly higher than in blood donors. No difference was seen in comparison with non-ulcer dyspepsia patients. The presence of Helicobacter pylori showed an inverse correlation with the extent of intestinal metaplasia. The intestinal type of carcinoma was associated with a higher bacterial load than the diffuse type. These data suggest that the presence of Helicobacter pylori in gastric mucosa could play a role in the pathogenesis of gastric carcinoma, especially in the young age group.     

HLA-DQA1 and -DQB1 genes and Helicobacter pylori infection in Italian patients with gastric adenocarcinoma

Both HLA-DQA1 and -DQB1 genes and Helicobacter pylori infection have been linked to gastric cancer. The aim of this work was to determine if HLA-DQA1 and -DQB1 alleles are presented at altered frequency in Italian patients with gastric adenocarcinoma and H. pylori infection. Oligotyping for HLA-DQA1 and -DQB1 and H. pylori serology was performed for 50 patients with gastric adenocarcinoma and compared with 80 patients with colonic adenocarcinoma and 179 healthy subjects. H. pylori infection was present in 76% of gastric cancer patients, 77.5% of colonic cancer patients, and 72% of controls. The prevalence of infection was not significantly different in the three groups of subjects sorted according to their HLA-DQA1 or -DQB1 status. Apart from HLA-DQA1* 0201, which was less common in patients with colonic carcinoma than controls, no other HLA-DQA1 and no HLA-DQB1 allele were present at altered frequency in patients with gastric or colonic cancer. Neither anatomical location and histological type of cancer nor the presence of lymph node or distant metastases were significantly associated with specific HLA-DQA1 or -DQB1 alleles or H. pylori infection. Both HLA-DQA1 and -DQB1 genes have a minor, if any, role in H. pylori infection and gastric carcinogenesis.     

Expression of cell cycle regulators and growth factor/receptor systems in gastric carcinoma in young adults: association with Helicobacter pylori infection

We studied the expression of cell cycle regulators and growth factor-receptor systems in gastric carcinoma in young adults and tried to clarify the specific alterations associated with H. pylori. We studied 33 young patients (18-29 years old, mean age 26.4) with gastric carcinoma. The patients were classified into two groups according to the degree of atrophic gastritis. Then we examined the expression of p53, cripto, cyclin-E, c-met, c-erbB2 and TGF-alpha immunohistochemically and compared the results between the two groups. The results were compared with 66 sex-, tumor histology-, and depth-matched elder controls (36-86 years old, mean age 64.0). H. pylori was judged by Giemsa staining. Seventeen patients had atrophic changes in the corpus (Group A), while 16 showed superficial gastritis or normal mucosa (Group S). All 17 patients of Group A showed H. pylori infection, while the 3 of the 16 members of Group S did not have H. pylori. p53 overexpression was observed more frequently in Group S (88%) than in Group A (41%, p<0.05). In the 3 patients without H. pylori infection, all carcinoma specimens showed p53 overexpression. Overexpression of cyclin-E was detected in 4 patients from Group S. On the other hand, cripto was observed more frequently in Group A than in Group S. No obvious differences were observed in c-erbB2, TGF-alpha and c-met expression. Overall, p53 overexpression was detected more frequently in younger than in older patients, whereas cripto expression was less detected. These results suggest that p53 and cyclin-E may act in an H. pylori-independent or -adjunctive manner for gastric carcinogenesis. Cripto expression might be correlated tightly with H. pylori infection.     

Decreased sensitivity of the ultrarapid urease test for diagnosing Helicobacter pylori in patients with chronic renal failure

Fifty patients with chronic renal failure and 50 asymptomatic healthy volunteers were studied to determine the prevalence of Helicobacter pylori in the two groups and to compare the results of ultrarapid urease test (URUT) and gastric histology for detecting H. pylori infection. Four gastric antral biopsy specimens were taken. Two specimens were used for the URUT and two were processed routinely. Sections were stained with H&E and Loeffler's methylene blue. Histological examination showed presence of H. pylori in 56% of patients with chronic renal failure (CRF) and in 78% of the controls. The difference was statistically significant (p < 0.05). The URUT was positive in only 16% of patients with CRF compared to 74% in the controls (p < 0.0001). The sensitivity and overall diagnostic accuracy of the URUT to diagnose H. pylori infection were 94.8 and 96%, respectively, in controls but were only 29 and 60%, respectively, in patients with CRF. It is concluded that the prevalence of H. pylori is significantly less in patients with CRF and that the URUT is less sensitive than gastric histology for diagnosing H. pylori infection in these patients.     

Helicobacter pylori test and medical reimbursement

There are an estimated 60 million people with Helicobacter pylori(H. pylori) infection who occupied 50% of the population of Japan. In Japanese medical reimbursement H. pylori tests were introduced on November 1, 2000 and they are able to use only to patients with gastric and duodenal ulcer. H. pylori tests were composed of rapid urease test, urea breath test, antibody test, bacterial culture and pathologic test. Payment of each test is 700 Yen. Classification and cost of H. pylori tests are shown. Usage of laboratory tests for H. pylori infection is mentioned. Those particular tests are useful to decrease the number of gastric and duodenal ulcer in Japan.     

Prevalence of H. pylori in patients with gastric cancer

The present study was taken with an aim to assess the prevalence of H. pylori in patients with gastric carcinoma and correlate it with gross appearance and histological type. Endoscopic biopsies from 54 patients with gastric carcinoma and 50 age and sex matched controls were taken after thorough upper gastrointestinal examination. Gross appearance of the tumour was noted and two biopsies each from the site of malignancy and from normal appearing areas were taken. Sections were stained with Haematoxylin & Eosin and Loeffler's methylene blue for histopathological details and presence of H. pylori. Prevalence of H. pylori in controls was slightly higher than the patients group (80% Vs 78%). Ulcerated type of gross appearance had maximum prevalence of H. pylori (88%). Prevalence of H. pylori was more in diffuse type of gastric cancer than intestinal type (86% Vs 68%). A significant association between H. pylori and grades of gastritis was noted (P < 0.01) in controls as well as in patient group but it failed to show a significant association with tumour grades, intestinal metaplasia, site of the tumour and age of the patients. So, it can be inferred that prevalence of H. pylori infection is not directly associated with pathogenesis of gastric cancer but it may act as a co-carcinogen by damaging the mucosa and thereby making it more susceptible to effects of carcinogen.     

Epstein-Barr virus (EBV) infection and p53 protein expression in gastric carcinoma

In the presented studies p53 protein expression was evaluated in samples of gastric carcinoma originating from 32 selected adult patients (with documented diagnosis of adenocarcinoma of the stomach and without the presence of Helicobacter pylori infection). Among the patients 14 individuals carried EBV-positive gastric carcinoma (group 1) while the 18 remaining patients carried EBV-negative gastric carcinoma (group 2). EBV infection was detected testing the tissue material for the presence of EBER by RNA in situ hybridization (ISH) and testing sera of the patients for EBV-specific antibodies. Expression of p53 protein was analysed using immunohistochemistry. Presence of p53 protein was noted in 9 (64.3%) cases of EBV-positive gastric cancer (group 1) and in 10 (55.5%) cases of EBV-negative gastric cancer (group 2). No significant differences were detected in the frequencies of p53 protein expression in the two studied groups. The results permit to conclude that abnormalities in p53 in gastric cancer are independent of EBV infection, even if EBV may participate in development of the tumour.     

Decreased gastric proliferation of foveolar epithelial cells after the eradication of Helicobacter pylori.

Increased epithelial cell proliferation is associated with an increased risk of gastric carcinoma. Helicobacter pylori infection is an established risk factor for gastric cancer and the organism has recently been classified as a group I carcinogen by an IARC working group. In this study, we describe differences in gastric epithelial cell proliferation between a H. pylori eradicated group (n = 21) and a not eradicated group (n = 8) after anti-H. pylori eradication therapy to show that increased cell proliferation is associated with H. pylori infection. H. pylori infection was determined by rapid urease test and immunohistochemical method with anti-H. pylori polyclonal antibody. Gastric epithelial cell proliferation was assessed using immunohistochemical method using Ki-67 monoclonal antibody. Ki-67 positive cells in H. pylori associated chronic active gastritis were observed in the glandular neck and the upper portion of foveolar epithelium. Patients who cleared their H. pylori infections showed a significant decrease of Ki-67 labeling index after therapy (0.73 +/- 0.10 vs. 0.48 +/- 0.08, p < 0.01). By contrast, Ki-67 labeling index before and after treatment in patients who remained positive for H. pylori showed no significant difference (0.78 +/- 0.08 vs 0.74 +/- 0.10, p > 0.05). These results indicate that H. pylori infection increases the proliferation of gastric foveolar epithelium, which is reduced by the eradication therapy. We suggest that anti-H. pylori eradication therapy can prevent mucosal cell proliferation to be closely associated with gastric carcinogenesis.     

Correlation analysis of riboflavin,RFT2 and Helicobater pylori in gastric carcinoma

Objective: To investigate the relationship between tissue riboflavin level and riboflavin transporter 2 (RFT2) protein expression, and the relationship between Helicobacter pylori (H.pylori) infection and the plasma riboflavin level in gastric carcinoma (GC). Methods: Enzyme-linked immunosorbent assay (ELISA) was used to detect tissue riboflavin level in patients with GC. Western blotting was applied to analyze the expression of RFT2 protein in 60 tissue samples from gastric carcinoma together with their normal tissues. The Warthin-starry method, rapid urease test and ¹⁴C-UBT were administered to detect the infection of H.pylori. High performance liquid chromatography (H.PYLORILC) was performed to detect plasma riboflavin level in the GC. Results: A significant decrease in the tissue riboflavin level was detected in GC samples compared to those in the normal mucous membrane (17.02±3.91 vs. 21.0±4.73; P = 0.043), and a significant decrease in the RFT2 protein was found in GC samples compared to those in the normal mucous membrane (0.92±0.39 vs. 1.23±0.51; P = 0.042). A positive correlation of tissue riboflavin level with defective expression of RFT2 protein was observed in GC patients (χ² = 1.969; P = 0.039). Plasma riboflavin level in gastric cancer without H.pylori infection group (1.6674 ng/mL ±0.37009 ng/mL) was higher than H.pylori infection group (1.2207 ng/mL ±0.17727 ng/mL, P = 0.043). Conclusion: The results indicate that RFT2 plays an important role in gastric carcinogenesis by modulating riboflavin absorption. H.pylori infection affects plasma riboflavin level and the prognosis of patients with gastric cancer.     

Argyrophilic nucleolar organizer regions in Helicobacter pylori-associated gastric lesions

Three hundred and fifty biopsies from patients undergoing upper gastrointestinal endoscopy were studied for histopathological changes, H. pylori infection and argyrophilic nucleolar organizer region (AgNOR) counts. Histopathological examination revealed normal gastric mucosa in 10 (2.85%), gastritis in 254 (72.56%), intestinal metaplasia in 12 (4.0%), dysplasia in 13 (3.7%) and adenocarcinoma in 61 (17.4%). The mean (SD) AgNOR count was 1.66 (0.20) in normal, 2.43 (0.64) in gastritis, 3.09 (0.52) in intestinal metaplasia, 4.17 (0.31) in dysplasia, and 6.57 (0.98) in carcinoma. A statistically significant difference was observed between the AgNOR count of normal gastric mucosa and gastritis (p<0.001), gastritis and dysplasia (p<0.001), and dysplasia and adenocarcinoma (p<0.001). A statistically significant increase in mean AgNOR count was found with increase in H. pylori density in gastric biopsies (p<0.001) with gastritis. No significant difference was observed between mean AgNOR count of intestinal and diffuse type carcinomas. The AgNOR count in gastric biopsies with adenocarcinoma and H. pylori infection was 7.03 (0.85) as compared to 6.89 (0.73) in gastric biopsies with evidence of adenocarcinoma but without H. pylori infection. The difference was not statistically significant. The findings support the role of H. pylori as a promoting agent in gastric carcinogenesis by stimulating gastric epithelial cell proliferation at the stage of chronic inflammation, thereby making the cells more susceptible to endogenous or exogenous carcinogenic agents.     

Helicobacter pylori infection and the risk of gastric carcinoma

BACKGROUND. Infection with Helicobacter pylori has been linked with chronic atrophic gastritis, an inflammatory precursor of gastric adenocarcinoma. In a nested case-control study, we explored whether H. pylori infection increases the risk of gastric carcinoma. METHODS. From a cohort of 128,992 persons followed since the mid-1960s at a health maintenance organization, 186 patients with gastric carcinoma were selected as case patients and were matched according to age, sex, and race with 186 control subjects without gastric carcinoma. Stored serum samples collected during the 1960s were tested for IgG antibodies to H. pylori by enzyme-linked immunosorbent assay. Data on cigarette use, blood group, ulcer disease, and gastric surgery were obtained from questionnaires administered at enrollment. Tissue sections and pathology reports were reviewed to confirm the histologic results. RESULTS. The mean time between serum collection and the diagnosis of gastric carcinoma was 14.2 years. Of the 109 patients with confirmed gastric adenocarcinoma (excluding tumors of the gastroesophageal junction), 84 percent had been infected previously with H. pylori, as compared with 61 percent of the matched control subjects (odds ratio, 3.6; 95 percent confidence interval, 1.8 to 7.3). Tumors of the gastroesophageal junction were not linked to H. pylori infection, nor were tumors in the gastric cardia. H. pylori was a particularly strong risk factor for stomach cancer in women (odds ratio, 18) and blacks (odds ratio, 9). A history of gastric surgery was independently associated with the development of cancer (odds ratio, 17; P = 0.03), but a history of peptic ulcer disease was negatively associated with subsequent gastric carcinoma (odds ratio, 0.2; P = 0.02). Neither blood group nor smoking history affected risk. CONCLUSIONS. Infection with H. pylori is associated with an increased risk of gastric adenocarcinoma and may be a cofactor in the pathogenesis of this malignant condition.