Patients and methods

Consecutive patients referred for upper gastrointestinal endoscopy, because of reflux complaints or dyspepsia, were eligible for inclusion. After informed consent, endoscopy was performed using the Olympus EVIS 100 video endoscope, and antral biopsy specimens were taken for detection of H pylori via culture and Gram stain, standard haematoxylin and eosin staining, rapid urease test, and immunoperoxidase staining, as described previously.⁸ In addition, a serum sample was taken and stored frozen at −70°C for the determination of IgG antibodies against H pylori and IgG anti-CagA antibodies.

Biopsy specimens were cut and stained according to the modified Giemsa protocol.⁹ Colonisation density of H pylori was assessed semiquantitatively at high power (magnification, ×400) in well oriented sections. The density was graded as follows: grade I, sporadic presence of bacteria within the mucus layer only detectable after scrutinised search of the entire biopsy specimen; grade II, clusters of bacteria present; grade III, bacteria covering at least half of the mucosal surface; and grade IV, the entire mucosal surface covered with bacteria. The topographical distribution of H pylori was described as follows: score A, H pylori closely attached to the mucosa; score B, H pylori attached to the mucosa and widely distributed in the gastric mucus; and score C, H pylori solely in the mucus.

Specific IgG antibodies against H pylori were measured in serum using an in house enzyme linked immunosorbent assay (ELISA) as described previously.¹⁰ An absorbance index above 0.32 was considered positive. Sensitivity and specificity of this ELISA for the detection of H pylori carriage were 98.5% and 91.7%, respectively.¹⁰

Determination of the H pylori cagA status was based on the presence of serum IgG antibodies to orv220, a 65 kDa recombinant CagA product purified from Escherichia coli.⁹ The presence of these antibodies was assessed by means of ELISA according to previously described methods.¹¹ The ELISA technique has been validated in the USA and yielded a sensitivity of 94.4% and a specificity of 92.5% for the detection of carriage of cagA⁺ H pylori strains. In addition, we validated the technique for the Dutch population by means of sera from 311 patients assessed to be H pylori negative by the combination of negative histology, culture, rapid urease test, and serology for H pylori. The mean result +2 SD in this population yielded an optical density cut off value of 0.458. All results above this value were considered positive.

Patients were considered to be H pylori positive if one of the histological or microbiological methods yielded a positive result. Only H pylori positive patients were included in our present study.

Previous use of acid suppressive treatment (H2 receptor antagonists or proton pump inhibitors) was assessed in all patients, but these patients were not excluded from the study.

Statistical analysis was done with χ² test for contingency tables and the t test. A p value below 0.05 was considered significant.

The study was approved by the medical ethics committee of De Heel Zaans Medisch Centrum.

Results

In total, 716 consecutive patients were studied; of these 345 were H pylori positive. From 293 of these patients (124 men, 169 women; mean age, 52.6 years; range 12–87), both serum and histology were available for the determination of IgG anti-CagA antibodies and modified Giemsa staining. There was no difference in age between men and women. CagA specific antibodies were present in 154 (52.5%) of the patients. There was no significant difference in colonisation density and cagA status (table 1). In the group of patients showing grade I bacterial density, a significantly higher number of patients received pretreatment with proton pump inhibitors. Table 2 shows the association between bacterial density and use of proton pump inhibitors or H2 blockers. When patients using proton pump inhibitors were excluded from the analysis no changes occurred in the relation between bacterial density and cagA status (table 3). Patients with grade I density were excluded from topographical scoring because H pylori was present only sporadically. Hence, correct topographic localisation in these cases was judged to be inadequate. There was no significant difference in topographical scores between cagA⁺ or cag⁻H pylori strains (table 4). Table 5 shows the relation between bacterial density and topographical score.

The mean (SD) IgG antibody titres were 0.71 (0.25) in patients with grade I H pylori density; 0.72 (0.26) in grade II; 0.74 (0.28) in grade III; and 0.68 (0.22) in grade IV. No significant differences were present. The mean (SD) anti-CagA titre was 0.84 (0.21) in patients with grade I bacterial density; 0.84 (0.21) in grade II; 0.89 (0.21) in grade III; and 0.73 (0.24) in grade IV. These results did not differ significantly. Exclusion of patients receiving treatment with proton pump inhibitors from the analysis had no effect.

Discussion

Helicobacter pylori colonisation causes inflammation of the gastric antrum and the corpus. The degree of inflammation,⁵ the density of H pylori, and the characteristics of the colonising strain correlate with each other.

The cag pathogenicity island is an important marker of virulence of H pylori. CagA⁺ strains are associated with increased intensity of gastric inflammation and increased mucosal concentrations of certain cytokines, particularly interleukin 8,¹² and are associated with the development of peptic ulcer and gastric cancer. The IgG antibody response correlates with the severity of gastritis.⁵ From this point of view, it could be argued that cagA⁺ H pylori strains elicit a higher immune response than cagA⁻ strains. In the literature, one report suggested that the density of H pylori in the gastric mucosa is higher in cagA⁺ than in cagA⁻ patients.⁷ The results of our study do not agree with this report. The obvious reason is that the results of quantitative culture, which can be judged as an indirect method, are different from direct visual analysis of biopsy specimens. In our patients no difference in bacterial colonisation density was noted when cagA⁺ and cagA⁻ H pylori strains were compared. Therefore, the presence of a more virulent strain does not necessarily result in gastric inflammation with a higher degree of bacterial colonisation density.

In our study, the cagA status of H pylori was assessed by the presence of antibodies against CagA, the protein encoded by the cagA gene. When these antibodies are present, it is assumed that a cagA⁺ H pylori strain is present.

It was recently reported that the CagA protein is delivered into gastric epithelial cells by H pylori, where it triggers profound changes in the morphology of the host cells.^2,3 The cytoskeleton is rearranged and a cup shaped pedestal forms beneath the bacterium. This process is thought to require intimate contact (attachment) between H pylori and the gastric cells. From these experiments it could be postulated that cagA⁺ H pylori strains are more closely attached to the epithelial cells; however, our present study showed that there was no difference in topographical localisation between cagA⁺ and cagA⁻ strains. The fact that topographical localisation is related to bacterial density suggests that topographical localisation is determined by colonisation density or vice versa, and not by the presence of the CagA protein. Finding a relation between density and colonisation is not unexpected because higher density might reflect an increased growth rate, whereas differences in localisation might reflect differences in the availability of nutrients.

It is well known that profound acid suppression with proton pump inhibitors results in a shift in H pylori from the antrum to the corpus, which is reflected by a decrease in bacterial density in the antrum.^13,14 In the Netherlands, many patients referred for upper gastrointestinal endoscopy have been pretreated with some type of acid suppressive drug. When all patients using proton pump inhibitors were excluded from analysis the results of our study did not change.

Although high IgG antibody titres indicate the presence of a cagA⁺ H pylori strain, this titre is of no value in determining the colonisation density of H pylori in the gastric antrum. In addition, the same is true for antibodies against CagA.

In conclusion, anti-CagA antibody titres do not correlate with antral colonisation density of H pylori and there is no difference in bacterial colonisation density between cagA⁺ and cag⁻ H pylori strains. The topographical distribution of H pylori is not determined by cagA status.