The seroprevalence of H. pylori ranges among Brazilians, and is dependent on age, socioeconomic status, and ethnicity, as well as sanitation conditions. High prevalence has been reported in the city of Fortaleza, with 73.3% positive cases in individuals 11-20 years old, and up to 87% in those over 60 years old 1. Notably, the incidence rate of gastric cancer is lower (16/100,000) in Fortaleza than in the city of São Paulo (22/100,000) http://www.inca.gov.br/, which has a lower prevalence of H. pylori (65.6%) 2. Furthermore, H. pylori seroprevalence was higher in African Brazilians compared to Caucasoids 2 and Japanese Brazilians 3, despite the higher mortality for gastric cancer among Japanese Brazilians compared to the indigenous population 4. These conflicting results suggest among other factors that more virulent H. pylori strains may be involved in gastric cancer outcome.

The cag (cytotoxin-associated gene) pathogenicity island (cagPAI) in H. pylori contains 31 putative genes 5 and encodes a type IV secretion system that delivers CagA into the cytosol, which is phosphorylated and activates phosphatase activity to initiate morphological changes of the cell, providing a potential mechanism by which chronic H. pylori infection may promote the development of gastric cancer 6. We previously showed that the cagPAI genes cagT, cagM, cagA and LEC (the left end of the cagII) were significantly associated with peptic ulcers 7, and these findings were later confirmed by another group in Brazil 8. The purpose of this study was to analyze the association of cagPAI genes and the blood group antigen binding adhesin babA of H. pylori with gastric cancer.

Patients were from the indigenous population of the city of São Paulo and were classified according to ethnicity as White, Brown (of White and African-Brazilian descent) and Black, none was Japanese-Brazilian. Among 68 consecutive gastric cancer patients, only 34 (11 intestinal-type, 23 diffuse-type) were H. pylori positive, 20 of which were men and 14 were women; the patients' mean age was 53.3 ± 12.4 years, and ranged from 24 to 75 years old; the median age was 52.5 years; 25 were White, 7 were Brown and 2 were Black.

Controls were selected among dyspeptic patients without previous history of peptic ulcer and gastric cancer, presenting a recent diagnosis of only gastritis (superficial or erosive) by upper gastrointestinal endoscopy, and positive for H. pylori infection. Of the 34 controls, 23 were women and 11 were men, with a mean age of 50.9 ± 9.8 years and a range from 32 to 70 years old; the median age was 48.5 years; 29 were White, 4 were Brown and 1 was Black.

The criteria for H. pylori positivity were the same for patients and controls, positive urease test and PCR with primers that amplify the species-specific 26-kDa antigen gene, using antrum and corpus biopsies 79. Patients and controls had not taken antimicrobials and acid suppression drugs for at least 30 days before the sample collection. In patients with gastric cancer, fragments were obtained from the normal appearance mucosa either by upper gastrointestinal endoscopy or after gastrectomy, immediately after opening the stomach. All patients provided informed written consent, and this study was approved by the local Ethics Committee.

DNA extraction from gastric biopsies of the positive urease tests and PCR were performed according to previously reported techniques. PCR analysis amplified regions of the cagPAI genes, babA, and DNA sequence of a 26-kDa species-specific protein antigen present in all strains of H. pylori 79. All the cases were confirmed H. pylori positive by urease test and by PCR for the species-specific antigen.

Fisher's exact probability test was determined using SPSS, and the odds ratio [OR] and 95% confidence interval [95% CI] were calculated using Microsoft Office Excel 2003. A value of P < 0.05 was considered statistically significant. The [OR] and [95% CI] are depicted in Table 1.

Comparison of the presence of cagPAI genes in H. pylori isolates from patients with gastric cancer and gastritis revealed that only two genes were significantly associated with gastric cancer: tnpA, detected in 73.5% (25/34) of the gastric cancer cases, with an odds ratio of 6.66 [95% CI, 2.30-19.25], and LEC, observed in 58.8% (20/34) of cases, with an odds ratio of 10.71 [95% CI, 3.07-37.28]. The other cagPAI genes analyzed in this study were detected equally in both groups. cagA was more frequently found in the gastric cancer group, though this difference was not significant; however, its presence may still be associated with gastric cancer, due to the odds ratio of 1.77 [CI 95%: 0.61-5.11]. The babA gene was more often found in H. pylori isolates from patients with gastritis (64.7%, 22/34) than in isolates from gastric cancer (41.2%, 14/34; p = 0.052), [OR, 0.38; 95% CI, 0.14-1.01]. The analysis of cagPAI and babA genes among the histological types (diffuse and intestinal) of gastric cancer, gender and the ethnic groups revealed no significant difference, data not shown.

As shown in Table 2, H. pylori isolates from gastric cancer cases were usually positive for both LEC and tnpA (p = 0.017 by Fisher's exact test). In Japan association of IS605 with other cagPAI genes, cag13 and cagA was previously reported in gastric cancer cases. However, as the role of this finding remains unclear, further study is necessary to determine its involvement in gastric cancer 10.

In this study, we analyzed the prevalence of H. pylori genes in patients from a Brazilian population with gastric cancer or gastritis. Unexpectedly, babA, the blood-group antigen binding adhesion targeting human Lewis^b surface epitopes on gastric epithelial cells associated with duodenal ulcer and gastric adenocarcinoma 11, was frequently deleted in gastric cancer H. pylori isolates (41.2%), compared to gastritis (64.7%); nonetheless, in the previously studied peptic ulcer group, 69.3% of H. pylori isolates also were babA positive 9. Thus, babA may be a frequent genotype with no particular role in the clinical outcome 12.

We previously showed that the cagPAI genes cagT, cagA, cagM and LEC were associated with peptic ulceration progression 7; this study revealed that only LEC and tnpA were associated with gastric cancer. In South Africa, LEC was frequently deleted in isolates from gastritis compared to those from gastric cancer and peptic ulcers 13. Nevertheless, LEC was not necessary for either the translocation of CagA that mediates intracellular disruption of growth regulation 14, or for the induction of IL-8 15, proinflammatory cytokine released upon H. pylori infection by gastric epithelial cells that induced expression and activation of epidermal growth factor receptor and proliferation 16. Our finding is consistent with a previous study reporting higher frequency of tnpA in Peruvian gastric cancer strains than in gastritis strains (9 of 14 versus 15 of 45, respectively; P = 0.04) 17.

We did not observe any significant association of cagA with cases of gastric cancer or gastritis, and this is consistent with previous reports in South Africa 13 and Germany 18, which failed to detect an association with gastric cancer.

Our study of a population in Brazil indicates an association with LEC and tnpA and gastric cancer; nonetheless, further studies are needed to confirm these results in a larger sample, as no specific role for these genes in gastric carcinogenesis has yet been identified.

cagPAI: cytotoxin-associated gene pathogenicity island; LEC: left end of cagII region

The authors declare that they have no competing interests.

RM participated in the conception, design, analysis and interpretation of data, statistical analysis, drafted the manuscript and revised final version; MSM carried out the molecular genetic studies; SBM and CLH obtained gastric biopsies; BZ provided gastric cancer patients and presented the study at the 8^th International Gastric Cancer Congress; FJC gave final approval of the version to be published. All authors read and approved the final manuscript.