Background

Helicobacter pylori is a spiral-shaped gram-negative urease-producing bacterium found in the gastric antrum and in areas of gastric metaplasia in the duodenum [1]. H. pylori infection is the cause of chronic gastritis and is important in the pathogenesis of peptic ulceration, gastric B cell lymphoma and gastric adenocarcinoma [1]. In Ghana, like many developing countries, the infection has a high prevalence rate (~80%) irrespective of the period of birth [2]. The incidence of Helicobacter pylori infection by rapid urease-campylobacter-like-organism (CLO)-testing in Ghanaian patients with dyspepsia referred for upper gastrointestinal endoscopy was found to be 75.4% [3].

Although infection is universally associated with gastritis, the development of clinical and endoscopic disease is dependent on a number of factors, including the virulence of the infecting strain, the susceptibility of the host and environmental co-factors [4]. The best understood bacterial virulence factor is the cytotoxin associated gene A (cagA) [5]. Its most hydrophilic region contains amino acid repeats including Glu-Pro-Ile-Tyr-Ala (EPIYA) [6]. In addition, the adjacent region of internal duplication has been shown to contain sequences derived from the duplication of three discrete segments of DNA [6, 7]. Helicobacter pylori strains possessing cagA induce more inflammation, ulceration and oncogenesis when compared with cag-negative strains [5]. Another such virulence factor is the vacuolating cytotoxin (vacA). The bacterial strains inducing more vacuolation, vacAs1 and m1, are closely associated with clinical disease [8].

Historically, observations postulated that gastro-duodenal disease was not common in Africa, however objective prospective endoscopic studies have demonstrated significant disease in the region [9]. Recent data also demonstrates the influence of Helicobacter pylori virulence factors on clinical and endoscopic disease in an endemic region like sub-Saharan Africa [10]. Here, for the first time we present data on characterization of H. pylori cagA and vacA genotypes and their association with gastro-duodenal disease in Ghanaian patients.

Methods

The aim of the study was to utilize a cross-sectional design to characterize H. pylori cagA and vacA genes in patients with dyspepsia at the Korle-Bu Teaching Hospital, the main tertiary referral medical centre in Accra serving the majority of the southern half of Ghana. A total of 159 patients referred with dyspepsia for endoscopy at the Korle-Bu Teaching Hospital, Accra were investigated for H. pylori between 2012 and 2014. Patients with prior H. pylori eradication treatment or proton-pump inhibitor-use 2 weeks were excluded from the study. All study participants had three gastric antral biopsies taken. Biopsies were obtained from 42 patients with erosive gastritis, 38 with duodenal ulcer (DU), 16 with endoscopic suspicion of gastric adenocarcinoma (GCA), 13 with both duodenal ulcer and erosive gastritis and 50 with non-ulcer dyspepsia (NUD). Each patient had H. pylori infection testing on antral biopsy using rapid-urease CLO examination at upper GI endoscopy (CLO testing kit: Cambridge Life Sciences Ltd, Cambridge, UK). The additional antral samples were stored at −20 °C in DNAgard until H. pylori vacA and cagA gene analysis (Biomatrica Inc, San Diego, USA).

H. pylori cagA and vacA genotype analysis

Genomic DNA was extracted from stored tissue samples collected from patients using QIAGEN DNeasy tissue kit, (Qiagen House, Crawley, West Sussex, UK). Detection of vacA and cagA genes was performed on gastric biopsy specimen by PCR as previously described by Rudi et al. [7].

Statistical analysis

Means were presented for continuous variables and frequencies for categorical variables. Chi square was used to demonstrate the differences between observed variables. Bonferroni correction was made for multiple hypothesis testing with a p-value of <0.0017 used to indicate statistical significance. Secondary analysis was done with logistic regression to determine presence or absence of endoscopic diagnoses with specific predictor variables using SPSS 16 software.

Results

During the study period, 159 patients with gastro-duodenal disease were screened for H. pylori with rapid-urease CLO examination, of which 113 (71.1%) were positive. The mean age was 49.8 years (SD 18.95). Fifty-three-percent were males and 47% females. The baseline characteristics of all patients by vacA and cagA genotype are shown in Table 1.

Table 1 Prevalence of H. pylori vacA and cagA genotypes in Ghanaian patients with gastro-duodenal disease
Full size table

In total, 110 (69.2%) vacAs1, 71 (44.7%) vacAm1, 35 (22.0%) vacAm2 and 119 (74.8%) cagA were detected. Eighty-two (74.5%) of 110 vacAs1 samples were cagA+, Fig. 1. Of the 119 cagA+ samples, 82 (68.9%) were found to have the vacAs1 genotype. The hydrophilic region and the region of internal duplication of cagA were detected in 77 (48.4%) and 109 (68.6%) respectively with 67 (42.1%) showing amplicons with both primer sets.

Fig. 1
figure 1

Illustrates amplicon size of 612 bp obtained for cagA gene-1/3 analysis. Ethidium bromide-stained 2.0% agarose gel electrophoregram of amplified cagA DNA fragments (612 bp) with primer cagA set 1/3. Lane M 100 bp molecular weight marker, Lanes 1–5 PCR positives, Lane N negative control

Full size image

In Table 2 the relationships between vacA, cagA genes and endoscopic diagnoses are presented. Duodenal ulcer was associated with the hydrophilic region of cagA (p = 0.002). By contrast, non-ulcer dyspepsia was associated with a lower prevalence of cagA, Table 2. In multivariate analysis of all patients, duodenal ulcer was more likely than other endoscopic diagnoses to have detectable cagA-(hydrophilic region) (OR 3.1 CI 1.2–7.9) or vacAs1m1 (OR 6.5 CI 1.2–34.0), Table 3.

Table 2 Relationship between vacA/cagA genotype and endoscopic diagnoses
Full size table
Table 3 H. pylori virulence factors associated with endoscopic diagnoses in multivariate analysis
Full size table

Discussion

In this cross-sectional study of the Ghanaian population with dyspepsia, we found a high prevalence of H. pylori, 71.1%, by rapid-urease CLO testing. Furthermore, the prevalence of H. pylori harboring the virulence factors, cagA and vacAs1, were found to be 74.8 and 69.2% respectively. The prevalence of vacAs1m1 was 25.2% while vacAs1m2 was 8.2%. Regarding the cagA gene, the hydrophilic region was more likely to be detected in duodenal ulcers while the region of internal duplication was associated with erosive gastritis, suggesting the effect of these genotypes in disease development. The significant association between cagA-(hydrophilic region) and duodenal ulcer persisted following multivariate analysis.

Majority of vacAs1+ samples were cagA + which was consistent with other studies [6, 11]. The expression of cagA gene is closely associated with that of vacuolating cytotoxin A (vacA) [11]. H. pylori cagA had a high prevalence in this study, 74.8%. This was demonstrated in other studies in Nigeria 90% [12], South Africa 95% [13]. Most H. pylori strains can be classified into two major groups. Type 1 have the gene coding for cagA and co-express cagA and vacA. Type 2 do not have the cagA gene and do not express cagA or vacA [14]. However, an intermediate phenotype has been identified expressing cagA independent of vacA or vice versa [14]. In Ghana the infecting Helicobacter pylori are seen to be of type 1.

The vacuolating cytotoxin A contains at least two variable regions, the signal (s) region which encodes the signal peptide and the middle (m) region [15]. The s region has two sub-types s1 and s2 while the m region has m1 and m2 [15]. The amount of cytotoxin produced is highest in the vacAs1m1 allele followed by the vacAs1m2 [16]. The frequency of vacAs1 and m1 vary across populations. The vacAs1 and m1 genes have been detected at a higher frequency in isolates from patients with DU in comparison with vacAs2 and m2, mostly in countries with a relatively low prevalence of H. pylori infection but also in South America and South Africa where the infection is endemic [16].

In our study population, vacAs1m1 genotype had a prevalence of 25.2%. Other African studies in Nigeria and Ethiopia had vacAs1m1 prevalence of 24 and 48% respectively [12]. Infection with H. pylori strains having the vacA s1m1 genotype (compared with s1m2 and s2m2) have also been associated with an increased risk of duodenal ulcer disease [17] as evident in this study. However, other reports from H. pylori endemic countries have not shown a significant association between vacAs1 or vacAs1m1 and gastro-duodenal disease [12]. A complex interplay of host genetic factors, environmental factors and other virulence factors of H. pylori are therefore important in determining the risk of gastro-duodenal disease [18].

CagA demonstrates considerable diversity in its 3′ region. Its most hydrophilic region contains amino acid repeats including Glu-Pro-Ile-Tyr-Ala (EPIYA) [6]. In addition, the adjacent region of internal duplication has been shown to contain sequences derived from the duplication of three discrete segments of DNA (D1, D2 and D3) [6, 7]. In this study, patients with duodenal ulcer were more likely to have detectable H. pylori cagA-(hydrophilic region) while H. pylori cagA-(internal duplication region) was associated with erosive gastritis. The PCR products of the region of internal duplications of cagA+ differ in size varying from 450 to 558 bps. H. pylori has been shown to express both gene regions in majority (98%) of patients [7]. However, in this study, cagA-(internal duplication region) was more prevalent than cagA-(hydrophilic region), (68.6% vs 48.4%) and 42.1% had both gene regions. This significant association between cagA-(hydrophilic region) and duodenal ulcer persisted following multivariate analysis.

Differences in cagA gene in our study population imply heterogeneity in the cag-pathogenicity island which may have an impact on clinical disease. This study however had limitations; PCR analysis was done on genomic DNA isolated from gastric biopsies which may have underestimated the vacA and cagA prevalence in the study population due to potential PCR inhibitors. Further evaluation would be required to characterize H. pylori bacterial diversity including cag-pathogenicity island gene polymorphisms and their impact on gastro-duodenal disease.

Conclusion

Majority of biopsies were colonized with H. pylori harboring both cagA and vacA. H. pylori cagA-(internal duplication region) was more prevalent than cagA-(hydrophilic region). Duodenal ulcer was more likely than other diagnoses to have detectable cagA-(hydrophilic region) or vacAs1m1.