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 [2,3,4]. 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 [4,5,6].

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 [7,8,9,10]. 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 [15,16,17].

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).

Results

Population study

Among 290 suspected non-cardia GC patients, 282 patients (225 males and 57 females; male/female, 3.9; mean age  ±  standard deviation, 62.5 ± 12.6 years old; age range, 29–87 years) were histologically confirmed as advanced non-cardia GC and were used for analysis. There were no differences in gender distribution by histological subtype (Table 1).

Table 1 The characteristics of gastric cancer population based on histologic features
Full size table

There was an increasing trend of intestinal type in the older age group; however, this trend did not reach statistical significance (P for trend = 0.081). In contrast, a converse tendency was observed for diffuse type (P for trend = 0.022). According to the Lauren classification, the mean age of individuals with intestinal type was significantly higher than that of patients with diffuse type (P = 0.027). The population study in Ho Chi Minh showed slightly significant lower age than that in Hanoi (mean age, 61.3 vs 64.5 years old, P = 0.045, respectively); however, among the diffuse type groups, there was no statistically different age distribution between Hanoi and Ho Chi Minh (mean age, 59.7 vs 63.2 years old, P = 0.13, respectively).

Based on Lauren’s criteria, the prevalence of patients diagnosed with intestinal type was significantly different between Hanoi and Ho Chi Minh (P = 0.017), but this was not seen with diffuse type or indeterminate type. Diffuse type cases were found to be predominant in Ho Chi Minh, whereas intestinal type cases were significantly more prevalent than diffuse and indeterminate types among patients in Hanoi (P = 0.036 and P = 0.021, respectively).

Prevalence of H. pylori infection

The positive H. pylori infection rate was 39.7, 42.9, 50.7, and 77.5% in the culture test, histological examination, CLO test, and serologic test, respectively. Given our criteria, in which at least one test yielded positive results, 224 (79.4%) subjects were judged to be infected with H. pylori. There was no significant difference in the prevalence of H. pylori infection between Hanoi and Ho Chi Minh or between male and female subjects (Table 2). We also failed to find a relationship between H. pylori infection and smoking or drinking habits. The infection rate in patients with diffuse type was significantly higher than those with intestinal type (84.6% vs 72.0%, P = 0.03) (Table 2). The significantly higher prevalence of infection seen in diffuse type than in intestinal type also was found in Ho Chi Minh (87.2% vs 69.1%, P = 0.01). The slightly higher infection rate in diffuse type was also observed in Hanoi (78.7% vs 76%, P = 0.9).

Table 2 Characteristics of the study population with Helicobacter pylori infection
Full size table

Limitation of serologic test is to detect both current and past infection. Taken together, in order to avoid the false positive results, H. pylori positive groups were divided into two groups; current infection and past infection. However, there was no association of age, sex, location, smoking, drinking as well as Lauren classification types between current and past infection (Table 2).

Distribution of cagA and vacA genotypes

A total of 112 strains could be isolated, and all strains possessed cagA (Table 3). With four distinct CagA EPIYA segments, East Asian-type CagA includes A, -B, and -D segments, while Western-type CagA includes A, -B, and -C segments [27]. Among isolated strains, 110 of 112 (98.2%) strains were East Asian-type, and most of these strains carried ABD type (93.8%), 4 strains (4.2%) had ABB’D, and 1 (1.0%) had BD (B’ is a subgroup of B). There were only 2 of 112 (1.8%) strains with Western-type CagA, and both carried ABC type. The carcinogenic potential of CagA is reported to be linked to its polymorphic EPIYA motif variants [29], and only polymorphisms of the EPIYA motif in B segments were found in this study: EPIYA (93.8%), EPIYT (2.7%) and ESIYA (3.6%). However, no polymorphism of the EPIYA motif in A and C segments was observed.

Table 3 Comparison of the characteristics of patients with Helicobacter pylori culture positive between gastric cancer and non-cancer group in Vietnam
Full size table

Regarding pre-EPIYA repeated region genotypes, 83.9% (94/112) of the strains possessed an 18-bp deletion type, which was reported as the Vietnamese-specific type in our previous study [28], and 11.6% (13/112) strains possessed the 39-bp deletion type, which is called the typical East Asian type. There were no deletion types typically observed in Western-type CagA strains. There was no difference in the pre-EPIYA deletion genotypes between the two cities (Table 3). More detailed analyses of the cagA strains are described in Additional file 1.

All strains possessed the vacA s1 genotype. The vacA m1 genotype was found predominantly in both cities (Table 3). When analyzing the combination of the vacA s and m regions, 77 (68.8%) strains were found to be s1m1, 35 (31.2%) were s1m2, and no strains with s2m1 or s2m2 were found. There were no differences between histological GC types and virulence factor genotypes (Table 4).

Table 4 The distribution of histological features based on virulence factor genotypes
Full size table

Comparison of characteristics between the gastric cancer and non-cancer groups among infected patients

In our previous study, we recruited 270 participants from Cho Ray Hospital, Ho Chi Minh and 108 Military Hospital, Hanoi [28]. Among them, 103 cases were H. pylori culture-positive. Based on endoscopic observation, 26 PUD and 77 gastritis cases were diagnosed. GC and mucosa-associated lymphoid tissue (MALT) lymphoma were not detected in the study. We considered the data from the previous study to be a non-cancer group.

The average age of the GC group was significantly higher than that of the non-cancer group [28] (mean age, 61.7 vs 44.6, P < 0.0001). The difference in age between the two groups was also observed in both cities (P < 0.0001). The percentage of male patients within the GC group was significantly higher than within the non-cancer group (75.9% vs 45.6%, P < 0.0001). The same trend of male prevalence between cancer and non-cancer groups was observed in Ho Chi Minh, but not in Hanoi (P < 0.0001 and P = 0.064, respectively).

The prevalence of H. pylori infection in the GC group was significantly higher than in the non-cancer group (79.4% vs 66%, P = 0.0003). Univariate analysis showed that the H. pylori infected patients had a higher risk for non-cardia GC than did non-infected patients (OR, 2.02; 95% CI 1.4–3.0; P = 0.0003).

Similarly to our current data, our previous study also showed that most H. pylori isolates from non-cancer patients possessed the East Asian-type cagA. There were no differences in pre-EPIYA repeated region genotypes between the two groups. Likewise, most strains isolated from both cancer and non-cancer groups were the vacA s1 genotype (Table 3). Only the vacA m genotype showed a difference between isolates from cancer and non-cancer groups: a higher prevalence of vacA m1 was found in the cancer group than in the non-cancer group (68.8% vs 44.9%; P = 0.001). Univariate analysis showed that vacA m1 conferred an increased risk for non-cardia GC (OR, 2.7; 95% CI 1.5–4.7; P = 0.001). In Ho Chi Minh, the prevalence of vacA m1 within the cancer group was also significantly higher than that of the non-cancer group (67.6% vs 31.9%, P < 0.0001). In Hanoi, although vacA m1 was more prevalent in the cancer group than in the non-cancer group (70.7% vs 56.9%), the difference did not reach statistical significance (P = 0.17). Between the two cities, the prevalence of vacA m1 in Hanoi was higher than that in Ho Chi Minh; however the statistically significant difference between these two cities was only observed in the non-cancer group, but not in the cancer group (56.9% vs 31.9%, P = 0.023; 70.7% vs 67.6%, P = 0.83, respectively) (Table 3). Univariate analysis indicated that vacA m1 conferred an increased risk for non-cardia GC in Ho Chi Minh (OR, 4.5; 95% CI 2.0–9.8, P < 0.0001), but not in Hanoi (OR, 1.8; 95% CI 0.8–4.4, P = 0.17).

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 [36,37,38]. 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.