Background
Helicobacter pylori infection is closely associated with the development of peptic ulcer disease and gastric cancer [
1‐
4].
H. pylori colonizes the gastric mucosa of 50% of the world’s population, with infection levels exceeding 70% in developing areas, such as Latin America and Africa [
5‐
10]. Several Southeast Asian countries also have a high prevalence of
H. pylori infection, particularly Thailand, where the
H. pylori infection rate ranges from 54.1%-76.1% [
11,
12]. Therefore, developing countries, including those located in Southeast Asia, are considered to have a higher incidence of
H. pylori-related diseases. However, the estimated age-standardized incidence rate of gastric cancer in Southeast Asian countries (10.2 and 4.7/100,000 in men and women, respectively) is lower than those in East Asia (e.g., 124.63/100,000 in Japan and 48.25/100,000 in Korea) and South America (e.g., 33.22/100,000 in Colombia) (
http://globocan.iarc.fr/). This phenomenon, which is characterized by high levels of
H. pylori infection, but a low incidence of gastric cancer, is referred to as the “Asian paradox” [
13,
14].
Recently, numerous studies have examined the relationships between
H. pylori virulence factors and outer membrane proteins with gastric mucosal inflammation and gastroduodenal disease development [
9,
10,
15,
16]. The
cag PAI including of
cagA encodes a putative type IV secretion system, which transfers a variety of multimolecular complexes, such as CagA, and across the bacterial membrane to the extracellular space or into other attached host cells [
17,
18].
H. pylori cagA plays important roles in gastric mucosal inflammation and injury in relation to activated inflammatory cells infiltration [
19,
20]. Activated neutrophils and mononuclear cells infiltrating into gastric mucosa with
H. pylori infection produce several pro-inflammatory cytokines (e.g., IL-1β, IL-6, IL-8 and TNF-α) and anti-inflammatory cytokines (e.g., IL-4 and IL-10). Greater than 90% of
H. pylori strains isolated from East Asian populations carries the
cagA gene [
7,
21]. In contrast, 40% of strains isolated in Western countries are
cagA-negative [
7,
21]. Moreover, as tyrosine phosphorylation of CagA occurs at EPIYA sites consisting of five amino acid residues, the structure and role of the
cagA EPIYA motif in
cagA gene have been investigated [
22‐
24]. The EPIYA motif exhibits genetic variation that occurs in four distinct segments, the EPIYA-A, -B, -C, and –D segments [
22‐
24]. The representative CagA of Western
H. pylori strains possesses a single EPIYA-A and EPIYA-B segment, followed by a 34-amino-acid EPIYA-C segment (EPIYA-ABC type). The C-terminal regions of East Asian and Western CagA are characterized by the presence of EPIYA-ABD and -ABC segments, respectively [
22‐
24]. In addition, we [
7] reported that
H. pylori strains isolated from East Asian and Western countries could be completely distinguished by polymerase chain reaction-based
cagA 5’ and 3’ region genotyping, and named the East Asian type
H. pylori strains as
cagA 1a type and the Western type
H. pylori strains as
cagA 2a type.
CagA 1a type is correspondent roughly to Western-type
cagA (EPIYA-ABC) genotype and 2a type to East Asian-type
cagA (EPIYA-ABD) genotype. However, it is unclear whether the cagA EPIYA motif is associated with the development of gastrointestinal disease in Southeast Asian populations.
Gastric epithelial cell injury associated with
H. pylori infections is caused by a vacuolating cytotoxin encoded by the
H. pylori vacA gene. The
vacA signal (s) region encodes the signal peptide and N-terminus of processed
vacA toxin. In
H. pylori, the
vacA s1 genotype is associated with fully active toxin, but type 2 genotype strains produce
vacA with a short N-terminal extension that blocks vacuole formation [
25]. The
vacA middle (m) region encodes part of the 55-KDa subunit located at the C-terminus and has two genotypes (m1 and m2); the former causes stronger vacuolating activities than the latter [
25]. Recently, a third polymorphic determinant of vacuolating activity located between the s- and m- regions was identified and termed the intermediate (i) region [
26]. In general, the
vacA s1, m1, and i1 genotypes of
H. pylori are associated with an increased risk of disease due to the enhanced production of toxin with markedly higher vacuolating activity than that of
vacA s2, m2, and i2 genotype strains, which are rarely associated with peptic ulcer disease and gastric cancer [
7,
8,
15,
25,
27‐
30]. However, it is unclear whether this association is observed in
H. pylori strains commonly found in Southeast Asian populations.
To date, 13 studies have investigated the
cagA status and genotyping of
cagA EPIYA motifs and
vacA s, m, and i-regions in
H. pylori strains found among Southeast Asian populations (Table
1) [
7,
31‐
42]. However, due to the small sample size in each report, it remains unclear whether
cagA-positive strains and the EPIYA motif and
vacA genotypes are associated with an increased risk for gastrointestinal disease in Southeast Asian populations [
7,
31‐
42]. Therefore, the present study was designed to elucidate the relationship between
H. pylori virulence-factor genotypes and
H. pylori-related disease susceptibility in
H. pylori-infected patients living in Southeast Asia.
Table 1
Reported
cagA
genotypes and
vacA
s, m, and i regions genotypes and
cagA
status in studies used in the present meta-analysis
Vietnam | | 2002 | 25 S
| 24 (96) | NA | NA | NA | 0 (0) | 24 (100) | 25 (100) | 0 (0) | 14 (56) | 11 (44) | NA | NA |
Vietnam | | 2009 | 103 S
| 98 (95) | 4 (4) | 0 (0) | 94 (96) | NA | NA | 103 (100) | 1 (1) | 44 (45) | 54 (55) | NA | NA |
Vietnam | | 2009 | 22 S
| 22 (100) | 1 (5) | 0 (0) | 21 (95) | NA | NA | NA | NA | NA | NA | NA | NA |
Vietnam | | 2010 | 100 | 95 (95) | NA | NA | NA | NA | NA | 100 (100) | 0 (0) | 48 (48) | 52 (52) | 94 (94) | 6 (6) |
Thailand | | 2002 | 8 S
| 8 (100) | NA | NA | NA | 4 (50) | 4 (50) | 8 (100) | 0 (0) | 6 (75) | 2 (25) | NA | NA |
Thailand | | 2004 | 98 S
| 98 (100) | NA | NA | NA | 50 (51) | 48 (49) | 98 (100) | 0 (0) | 71 (72) | 27 (28) | NA | NA |
Thailand | | 2007 | 135 | 119 (88) | NA | NA | NA | NA | NA | 132 (100) | 0 (0) | 73 (63) | 42 (37) | NA | NA |
Thailand | | 2008 | 112 S
| 110 (98) | NA | NA | NA | NA | NA | 112 (100) | 0 (0) | 65 (58) | 47 (42) | NA | NA |
Malaysia | | 2005 | 127 S
| 107 (84) | NA | NA | NA | NA | NA | 117 (92) | 10 (8) | 81 (64) | 46 (36) | NA | NA |
Malaysia | | 2006 | 73 S
| 58 (79) | NA | NA | NA | NA | NA | 65 (89) | 8 (11) | 46 (63) | 27 (37) | NA | NA |
Malaysia | | 2009 | 93 | NA | 23 (22) | 13 (12) | 70 (66) | NA | NA | NA | NA | NA | NA | NA | NA |
Malaysia | | 2009 | 126 S
| NA | 34 (28) | 19 (16) | 68 (56) | NA | NA | NA | NA | NA | NA | NA | NA |
Malaysia | | 2010 | 159 S
| 159 (100) | 71 (45) | 0 (0) | 88 (55) | NA | NA | 155 (97) | 4 (3) | 94 (59) | 65 (41) | 154 (97) | 5 (3) |
Singapore | | 2000 | 108 S
| 95 (88) | NA | NA | NA | NA | NA | 103 (99) | 1 (1) | 39 (38) | 65(62) | NA | NA |
Methods
Study selection
A literature search was performed using the PubMed databases for articles written in English and published before June 2011. The following search words were used: 1) cagA or EPIYA, 2) vacA or vacuolating cytotoxin, 3) pylori or Helicobacter, and 4) genotype. We did not include abstracts or unpublished articles. We conducted a combined analysis to determine the prevalence of cagA-positive strains, cagA EPIYA motif genotypes, and vacA genotypes in H. pylori strains found in Southeast Asia and their association with gastrointestinal diseases.
Inclusion Criteria
The following criteria were applied to select published case–control studies examining the relationship between cagA EPIYA motif genotype or vacA genotype and clinical outcomes in adult populations infected with H. pylori isolated in four Southeast Asian countries (Vietnam, Thailand, Singapore, and Malaysia): the presence and genotypes of the cagA EPIYA motif (ABC, ABCC, and ABD genotypes) and vacA (vacA s-, m-, and i- regions) were examined by polymerase chain reaction (PCR) and original articles published in English. The references cited in these manuscripts were also screened using the same inclusion criteria. When it appeared that the same subjects were included in multiple reports, only the earliest article was selected.
Data analysis
As several studies did not measure all parameters simultaneously (cagA status and vacA s-, m-, and i-region genotypes), the patient and strain numbers (H. pylori genotype number) did not match in the following analyses.
Statistical differences in the prevalence of cagA status, EPIYA motif genotype, and vacA genotype among the individual countries and ethnic groups were determined by the chi-squared test and Fisher's exact test. The effects of cagA status, EPIYA motif genotype, and vacA genotypes on the risk of gastric cancer and peptic ulcer disease were expressed as odds ratios (ORs) with 95% confidence intervals (CIs) with reference to non-ulcer dyspepsia (NUD) subjects infected with H. pylori. NUD was defined as endoscopical gastritis with no peptic ulcer disease or gastric cancer. In this study, patients with ‘NUD’ were regarded as the control group. All p values were two-sided, and p values <0.05 were considered statistically significant. Calculations were carried out using the statistical software StatView 5.0 (SAS Institute, Cary, NC, USA). Meta-analyses were performed using Comprehensive Meta-Analysis software (version 2, Biostat, Englewood, NJ).
Discussion
Highly virulent genotypes of
H. pylori are associated with gastric epithelial damage, including gastric mucosal atrophy, in infected patients [
30]. Notably, the
vacA s1, m1, and i1 genotypes and
cagA-positive strains of
H. pylori are linked to elevated inflammatory cell infiltration compared to that induced by
vacA s2, m2, and i2 genotypes and
cagA-negative strains [
48,
49]. In the present study, our meta-analyses revealed a significant relationship between
H. pylori virulence factor-associated genotypes, particularly
vacA m-region genotype and
cagA status, and an increased risk for the development of peptic ulcer disease in Southeast Asian populations.
An important characteristic of Southeast Asian countries is that several geographic populations live together. Recently, Breurec et al. [
50] reported that five major types of historical human migration patterns have occurred in Southeast Asia: i) migration from India introducing hpEurope bacteria into Thailand, Cambodia, and Malaysia; ii) migration of the ancestors of Austro-Asiatic speaking people carrying hspEAsia bacteria into Vietnam and Cambodia; and iii) migration of the ancestors of the Thai people from Southern China into Thailand carrying
H. pylori of population hpAsia2; iv) migration of Chinese to Thailand and Malaysia within the last 200 years resulting in the spread of hspEAsia strains; and v) migration of Indians to Malaysia within the last 200 years distributing both hpAsia2 and hpEurope bacteria. Therefore, both Western and East Asian strains according to EPIYA motif genotyping can be observed in Southeast Asia. East Asian-type strain (EPIYA-ABD) was associated with an approximately three-fold increased risk of disease. Therefore, we concluded that the risk of developing
H. pylori-associated diseases varies among different geographic populations despite living simultaneously in the same country.
The “Asian paradox” might be explained by the widespread prevalence of weakly cytotoxic strains and correspondingly low frequency of
H. pylori-associated diseases. However, here, the prevalence of strains with the
vacA s1 region among Southeast Asia populations was similar to that of East Asian and Latin American populations, who have a higher risk of gastric carcinogenesis [
7,
9,
10]; half of the isolated
H. pylori strains were Western-type strains with an EPIYA-ABC genotype. The sequence of the EPIYA-D type perfectly matches the high-affinity binding sequence for SHP2 domains of SHP-2 (pY-[V/T/A/I/S]-X-[L/I/V]-X-[F/W]) [
51,
52]. In contrast, the sequences of the EPIYA-C type differ from the SHP-2 binding sequence by a single amino acid located in the pY + 5th position [
51,
52]. As a result, East Asian-type CagA exhibits stronger binding activity for SHP-2 and a greater ability to induce morphological changes in epithelial cells than the Western type. The estimated age-standardized incidence rates of gastric cancer in Vietnamese patients (24.4 and 14.6/100,000 in men and women, respectively) infected with East Asian-strains of
H. pylori are relatively higher than those in Thailand (4.2 and 3.0/100,000 in men and women, respectively) and Malaysia (10.7 and 6.4/100,000), where the majority of the population is Thai and Malay, and are infected with Western-type strains (
http://globocan.iarc.fr/). Therefore, the lower prevalence of higher virulence strains (e.g., ABD type) may explain the low frequency of gastric cancer observed in Southeast Asia, particularly in the southern part.
In Vietnam,
H. pylori infection was detected at a rate of approximately 66% and was shown to be strongly associated with active gastritis, atrophy and intestinal metaplasia [
33]. As genetic characteristics of
H. pylori in Vietnam, the majority of isolates possess
cagA,
oipA
“on
”,
vacA s1, and
vacA i1 genes, while the incidence of
vacA m1 gene is less frequent [
7,
31,
33]. Here, the prevalence of the
vacA m1 genotype in patients residing in Hanoi and Ho Chi Minh were 58% and 34%, respectively. The prevalence of peptic ulcer disease in Hanoi, located in the northern part of Vietnam, is significantly higher than that in the southern city of Ho Chi Minh, despite similarities of ethnicity and diet [
33]. Moreover, the age-standardized incidence rate of gastric cancer in Hanoi was approximately 1.5-fold higher than that in Ho Chi Minh (27.0 vs. 18.7 cases per 100,000 males, respectively) [
33]. The higher prevalence of peptic ulcer disease and gastric cancer observed in Hanoi might be attributable to the higher prevalence of
H. pylori strains carrying
vacA m1 in this region. Although Vietnam is located between regions with a high and low risk of gastric cancer, the rate of gastric cancer in Vietnam is approximately three times lower than that in Japan and Korea, a finding that might also explain the low prevalence of
H. pylori strains carrying
vacA m1 (58% vs. nearly 100% in East Asia) [
7,
16,
53].
Thailand has a population of 60 million people and is comprised of two major ethnic groups: Thai and Chinese. Eight types of
H. pylori strains were identified in Thailand; half of the strains possessed genotypes typically found in South Asia (
cagA type 2a/cag right-end junction type III and
vacA s1a-m1c), while the other half consisted of genotypes typically encountered in East Asia (
cagA type 1a/cag right-end junction type II and
vacA s1c-m1b) based on combinations of
cagA EPIYA motif genotype, genotype at the right end of the
cag pathogenicity island into five subtypes according to deletion, insertion and substitution motifs (e.g., Type II strains were predominant in China and Japan and type III strains were most common in India and type III strains were typical of strains from India) and
vacA s-region genotype [
7,
54]. Thailand was considered to be a cross roads with respect to
H. pylori genotypes, as isolates from ethnic Thai were commonly either South/Central Asian or mixed genotypes (East Asian and South/Central Asian) based on the
cagA,
cagPAI, and
vacA genes, whereas isolates from ethnic Chinese were typically the East-Asian genotype [
55]. Although the incidence of gastric cancer in Thailand is lower than that of East-Asian populations, the incidence of gastric cancer is not rare among Chinese living in Thailand, as the majority of gastric cancers (82%) occurred among ethnic Chinese or Thai-Chinese. Therefore, the incidence of gastric cancer varies among different geographic populations, and the high prevalence of higher-virulence strains may be associated with the high frequency of gastric cancer in Chinese residing in Thailand, which is typically a low-incidence area.
In Malaysia, three distinct ethnic groups predominate: Malays, Chinese, and Indians. Chinese and Indians have migrated to Malaysia for nearly three consecutive generations. The
vacA s1c genotype was also the predominant genotype detected among the Chinese patients residing in Malaysia, while s1a was predominant in Indians and Malays in Kuala Lumpur [
37].
The present study has some limitations. First, in this study, because the prevalence of
cagA-negative strains was very low in Southeast Asian countries, we did not demonstrate the relationship between risk of
H. pylori-related diseases and
cagA-negative strains, which are 20-80% of
H. pylori strains isolated from US and European population. Second, there was no information about biochemical role among different
H. pylori virulent factors (e.g.,
cagA status,
cagA EPIYA motif and
vacA genotypes) in pathogenesis for peptic ulcer disease and gastric cancer development [
56]. Recently, Mueller, et al. [
56] reported association of c-Src and c-Abl kinases and phosphorylation of EPIYA motif, and differences of phosphorylation between East Asian-type strain (EPIYA-ABD) and Western type strain (EPIYA-ABC). Therefore, further studies will be required to clarify the role of
H. pylori cagA EPIYA motif and
vacA genotype for the development of gastrointestinal diseases.
Conclusion
We demonstrated that the prevalence of specific
vacA m-region and
cagA EPIYA motif genotypes was found to vary significantly among the respective Southeast Asian countries. Moreover, our present meta-analyses identified a significant relationship between
vacA m-region genotype and
cagA status and the development of diseases in Southeast Asian. Importantly, most of the
H. pylori strains isolated from countries with high incidences of gastric cancer concurrently possessed virulent genotypes such as
vacA s1/m1 and East Asian-type
cagA[
7]. In contrast, in countries with a low gastric cancer incidence, such as Thailand and Malaysia, a considerable proportion of
H. pylori isolates exhibited less virulent genotypes, such as
vacA m2 and Western-type
cagA[
7]. Based on the age-standardized incidence rate of gastric cancer, Asian countries can be categorized as either high-risk (e.g., Japan, Korea, and China), intermediate-risk (e.g., Vietnam), or low-risk (e.g., Thailand and Indonesia). The clinical usefulness of
cagA EPIYA motif and
vacA genotype testing must be evaluated in studies using a large number of individuals with
cagA EPIYA motif and
vacA genotypes.
Competing interests
The authors declare no competing interests related to this study.