Background
Gastric cancer (GC) is the fifth most common cancer worldwide and the third most common cancer in Asia (GLOBOCAN 2012;
http://globocan.iarc.fr). Although GC incidence and mortality rates have been slowly declining over the last few decades in most countries, GC still remains a significant public health problem [
1]. In Asia, where GC is the third most common cause of cancer death, the incidence and mortality of GC vary widely among countries, even among different ethnic groups or regions in the same country [
2‐
4]. Etiologically, GC is closely associated with many factors such as age, dietary and lifestyle factors, genetics, and especially
Helicobacter pylori (
H. pylori) infection and its virulence factors [
4‐
6].
Helicobacter pylori is a spiral, gram-negative bacterium that has colonized the stomachs of approximately half of the world’s population.
H. pylori infection is the strongest risk factor for GC, and most
H. pylori-infected people develop chronic gastritis [
7‐
10]. However, only a small proportion of infected subjects has a potentially higher risk of developing non-cardia or distal GC (antrum or corpus) [
9]. One reason underlying this observation is the difference in bacterial pathogenicity.
H. pylori strains are genetically highly diverse, and several genes/genotypes are strongly associated with virulence and are consequently linked to carcinogenicity. Among virulence factors, the diverse clinical outcomes of
H. pylori infection are strongly associated with the diversity of
cagA and
vacA [
8,
11].
CagA is the most widely studied
H. pylori virulence factor [
12]. More than 90% of
H. pylori isolated from East Asian countries carry
cagA, whereas only 50–70% of
H. pylori isolated from Western countries do [
13,
14]. The CagA protein has been classified into Western-type and East Asian-type based on the sequences of repeat regions of CagA containing Glu-Pro-Ile-Tyr-Ala (EPIYA) motifs [
12,
13]. Individuals infected with East Asian-type CagA
H. pylori have been reported to have an increased risk of peptic ulcer disease (PUD) and/or GC, compared to those infected with Western-type CagA strains [
15‐
17].
VacA is another extensively studied
H. pylori virulence factor [
12]. The vacuolating activity varies among different
H. pylori strains [
18] because of the different structure of
vacA, which consists of the signal (s) (s1 and s2) and the middle (m) (m1 and m2) regions [
19]. It has been showed that
vacA s1m1 exhibits the most cytotoxicity, followed by
vacA s1m2, whereas
vacA s2m2 does not display any cytotoxicity [
19]. Furthermore, individuals infected with
vacA s1 or m1
H. pylori have an increased risk of PUD and/or GC compared to those infected with s2 or m2 strains [
19,
20].
Vietnam has emerged as a country with the highest age-standardized incidence rate (ASR) of GC (16.3 cases/100,000 for both sexes) in Southeast Asia (GLOBOCAN 2012;
http://globocan.iarc.fr). In keeping with the consensus, our previous studies have reported the high prevalence of
H. pylori infection in Vietnam and its strong association with PUD, active gastritis, atrophy, and intestinal metaplasia [
21]. However, these studies were from a short-term, cross-sectional design study, which is why the samples included only gastritis and PUD. Therefore, in this study, we aimed to exclusively study non-cardia GC subjects in order to describe the association between
H. pylori infection and non-cardia GC, as well as characteristics about demographics and histopathology of non-cardia GC in Vietnam. In addition, statistical data indicate that the ASR of GC in Hanoi, the capital city located in the north, is approximately 1.5 times higher than that in Ho Chi Minh, the second largest city located in the south. This phenomenon was called the “Vietnam enigma” [
22]. In an effort to clarify this phenomenon, our previous study included gastritis and PUD patients, but not GC patients, and showed that the
vacA m1 type was associated with an increased risk for PUD and was significantly more common in Hanoi (54%) than in Ho Chi Minh (31%) [
21]. We, therefore, suggested that the
vacA m1 type might contribute to the difference in the incidence of GC between these two cities.
Methods
Patients and H. pylori
Suspected GC patients undergoing gastroscopy at the two endoscopy centers—Cho Ray Hospital, Ho Chi Minh and 108 Military Hospital, Hanoi from November 2012 to May 2014—were recruited. Exclusion criteria included a history of gastrectomy; previous H. pylori eradication therapy; treatment with bismuth-containing compounds, H2-receptor blockers, or proton pump inhibitors within 2 weeks of the start of the study; and patients with suspected recurrent Billroth anastomosis cancer, entire stomach tumor, or cardiac tumor. Local ethics approval was obtained from the Ethics Committee of Cho Ray Hospital and 108 Military Hospital, and written informed consent was obtained from all patients. The study was also approved by the Ethics Committee of Oita University Faculty of Medicine, Japan.
Before endoscopy, patients were interviewed by trained medical staff to obtain their medical history and lifestyle factors. Depending on the location of the tumor, at least eight biopsy specimens (three from the normal-looking mucosa at least 2–3 cm apart from the tumor margin in the antrum, one from the upper posterior or anterior wall of the corpus, and four from the tumor margin) were taken. Three specimens from the antrum were used for
H. pylori culture, a rapid urease test, and histological examination. When the antrum was occupied by the tumor, the specimens were obtained from the lower corpus. The corpus specimen was used for histological examination. Suspected GC cases were confirmed by histological examination. GC was classified into intestinal or diffuse type according to the Lauren classification [
23]. Blood samples from all participants were collected on the same day. Serum was separated and frozen at −80 °C until analysis.
Determination of H. pylori status
To maximize the diagnostic accuracy, four different tests were used for the diagnosis of H. pylori infection: culture, histology including immunohistochemistry (IHC), a rapid urease test (CLO test, Campylobacter-like organism test, Kimberly–Clark Ballard Medical Products, Roswell, GA, USA), and serum H. pylori antibody presence.
Helicobacter pylori culture was performed as previously described [
21]. For histology, hematoxylin and eosin (HE), Giemsa staining, and IHC with an anti-
H. pylori polyclonal antibody (Dako, Denmark) were performed, and the updated Sydney system was used, as described previously [
24,
25]. For the serologic test, we used an ELISA kit (Eiken Co., Ltd, Tokyo, Japan) to detect anti-
H. pylori antibody according to the manufacturer’s instructions.
Helicobacter pylori-positive status was determined by at least one positive result among four tests. H. pylori-negative status was diagnosed if all four tests gave negative results.
Determination of cagA and vacA virulence factors
cagA status was determined by PCR for the conserved region and by direct sequencing, as described previously [
26]. Briefly,
cagA was amplified by using the primers cagOMF (5′-AGC AAA AAG CGA CCT TGA AA-3′) and cagOMR (5′-AGT GGC TCA AGC TGC TGA AT-3′). The PCR conditions were initial denaturation for 5 min at 94 °C, 35 cycles of amplification step (94 °C for 30 s, 56 °C for 30 s, 72 °C for 30 s), and a final step of 72 °C for 7 min. The amplified fragment was detected by a 1.5% agarose gel electrophoresis using an UV transilluminator. The PCR products were purified using a QIA quick purification kit (Qiagen) and sequenced by using an AB 3130 genetic analyzer (Applied Biosystems, Foster City, CA). The CagA types (East Asian-type CagA and Western-type CagA) were defined according to the flanking sequences of EPIYA motif patterns. Taken together, the strains possessed EPIYA-A, EPIYA-B, and EPIYA-C segments was considered as Western-type CagA and the strains possessed EPIYA-A, EPIYA-B and EPIYA-D segments was considered as East-Asian-type CagA.
vacA genotyping (s1, s2, m1, and m2) was performed as described previously [
19,
27]. Briefly,
vacA s region was amplified by using the primers VA1-F (5′-ATG GAA ATA CAA CAA ACA CAC-3′) and VA1-R (5′-CTG CTT GAA TGC GCC AAA C-3′); and
vacA m region was amplified by using the primers VAG-F (5′-CAA TCT GTC CAA TCA AGC GAG-3′) and VAG-R (5′-GCG TCA AAA TAA TTC CAA GG-3′). The amplified fragment was detected by using a 2% agarose gel electrophoresis using an UV transilluminator. Primers for s region yielded a fragment of 259 bp for s1 variants and 286 bp for s2 variants. Primers for m region yielded a fragment of 570 bp for m1 variants and 645 bp for m2 variants. As a control, the data for the
cagA and
vacA genotypes of gastritis and PUD patients in our previous study were compared with data in the present study [
28].
Statistical analysis
Chi squared tests, Fisher’s exact tests, independent sample t tests, and one-way ANOVA tests were used. Differences at P < 0.05 were regarded as statistically significant. Data analysis was performed using SPSS statistical software v.20.0 (SPSS Inc., Chicago, USA).
Discussion
In Vietnam, GC remains the fourth most common type of cancer and is the third leading cause of cancer-related death in both genders (
http://globocan.iarc.fr); therefore, GC is still an aggressive disease that continues to have a daunting impact on public health. This is the first study to describe the characteristics of the demographics and histopathology of non-cardia GC in the Vietnamese population. Consistent with many previous studies [
30], our data in Vietnam showed that non-cardia GC has a late age of onset and a male preponderance. The incidence of non-cardia GC increases with age, with the peak mainly occurring after 50 years. These characteristics may reflect the underlying variation in
H. pylori and lifestyle and environmental exposures.
Development of GC is a multistep and multifactorial process. The intestinal type is often related to
H. pylori infection, diet, and lifestyle, while the diffuse type is more often associated with genetic abnormalities [
31]. In this study, the overall proportion of diffuse carcinoma was found to be more predominant than the intestinal type, especially among younger patients and
H. pylori-positive subjects. These results in Vietnam, which is a country with an intermediate risk of GC, are consistent with the global tendency, in which the intestinal type tends to predominate in geographic regions with a high incidence of GC and is less likely to be found in areas where the frequency is declining [
32]. Moreover, all patients in this study were in the advanced stage of GC. This might be attributed to vague and non-specific symptoms in the early stages of GC and might cause the relatively poor prognosis of GC patients in Vietnam. In contrast, in Japan, due to the well-established strategy for GC prevention screening, most of the new cases are now diagnosed at an early stage. There, the patients prognosis is extremely good, with more than 90% surviving for 5 years or more [
33]. With diffuse type preponderance in the study population, it reflects a younger trend in developing GC. Therefore, there is an urgent need to establish a screening guideline for GC, especially early GC in Vietnam.
The etiology of GC is multifactorial; however,
H. pylori is attributable to about 70% in GC [
34]. In this study, the serologic test showed the highest prevalence of
H. pylori infection. Culture and histology were considered to be standard methods for detecting
H. pylori because of its direct visibility. However, the two methods showed the lowest prevalence of infection. There are some possible explanations for this difference: first,
H. pylori does not colonize areas affected by cancer, intestinal metaplasia, or atrophy, especially in advanced GC; secondly, there may be a bias of biopsy sampling areas. The advantage of serology is that both current and past infections can be detected. This means that even if
H. pylori is lost from the stomach during the development of advanced GC, we can still obtain evidence of previous colonization of the organism for years. Although patients with history of
H. pylori eradication were excluded, to avoid the bias of detection methodologies, we also analyzed current infection (at least two test positive) and past infection (only positive with serology test). However, there was no differences on age, sex, location, smoking, drinking, as well as Lauren classification types between these two groups. This is the reason why we did not apply the same criteria used in the previous study;
H. pylori-positive status was determined by positive culture or, in the case of negative culture, by at least two positive results among remaining tests. Based on our current criteria, the prevalence of
H. pylori infection was significantly higher in the cancer group than in the non-cancer group (79.4% vs 66%). In concordance with meta-analyses [
35], our data showed that
H. pylori infection was associated with approximately twofold increased risk of developing non-cardia GC. Preventative measures for GC have been developed with the focus on
H. pylori therapy, and this has succeeded in decreasing the incidence as well as the mortality of GC [
36‐
38]. The Asian Pacific gastric cancer consensus has recommended screening and eradication of
H. pylori in communities with a high risk of GC (ASR > 20/100.000). The ASR of Vietnam is 16.3/100.000, which is near the standard threshold. With the burden of GC in Vietnam, an
H. pylori eradication-based GC prevention strategy should be cost-effective and plausible.
Helicobacter pylori has emerged as the most important causal factor for GC; however,
H. pylori infection alone is insufficient to cause GC. One hypothesis is that not all
H. pylori strains are equal in virulence. Despite the similarity in ethnicity and dietary factors, the incidence of GC in Hanoi is 1.5 times higher than in Ho Chi Minh. In efforts to clarify this phenomenon, our previous study also partly proved the outstanding importance of virulence factors over the presence of bacteria in describing GC incidence in Vietnam [
21]. Our results indicated that
vacA m1, but not infection rate, correlated with an increased risk for PUD and contributed to the difference in the prevalence of GC between Hanoi and Ho Chi Minh.
Like several countries in East Asia where most strains harbored East Asian-type CagA regardless of the disease [
39], the CagA presence failed to explain the differences in the spectrum of diseases of gastroenterology, including the CagA full-length sequences analysis (Additional file
1). In contrast, our previous study indicated that the occurrence of
vacA m1 genotype of the non-cancer group in Hanoi was significantly higher than that in Ho Chi Minh (56.9% vs 31.9%) [
21]. Nevertheless, the study only showed indirect evidence because of the lack of a comparison between GC and control cases. As a satisfactory answer to this concern, we first confirmed that there was a difference in the prevalence of
vacA m sub-types between GC and non-cancer patients. Our current analysis showed that the occurrence of the
vacA m1 genotype was significantly higher in the in the GC group than in the non-cancer group (68.8% vs 44.9%), and the significant tendency was still observed in Ho Chi Minh (67.6% vs 31.9%). It has been accepted that
vacA m1 is more toxic than
vacA m2, and individuals infected with
vacA m1 strains have an increased risk of PUD and/or GC, compared to those infected with
vacA m2 [
19,
20]; therefore, our study strongly indicated that
vacA m1 play an important role in GC in Vietnam, especially in Ho Chi Minh.
A limitation of this study is the difficulty of conducting a strong analytic, cross-sectional study. Because the number of cases of non-cardia GC is low, we conducted a long-term project for non-cardia GC subjects and compared those results with our previous study. Some information gathered in these two studies was not synchronous, which can cause some biases during interpretation. However, this study provides stronger evidence by using GC subjects.
In conclusion, we first described the characteristics of non-cardia GC in Vietnam late-age onset, predominance in males, and diffuse type. H. pylori infection was common and associated with the development of non-cardia GC. We also confirmed the association of the vacA m1 genotype with an increased risk of developing non-cardia GC, which might contribute to the difference in the incidence of non-cardia GC between Hanoi and Ho Chi Minh.
Authors’ contributions
TTB, HDQD, VVK, YY conceived and designed the study; VPT, HDQD, PHT, TDT, PQH, VVK contributed by collecting samples; TTB, TTHT performed the experiments; TTB, VPT, TTHT, SR and YY contributed to analysis and interpretation; TTB, VPT, TTHT and YY drafted the manuscript. All authors read and approved the final manuscript.