Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

1 Introduction

Although gastric cancer rates appear to be declining worldwide, it is still a huge problem. It was estimated as the fourth most common cancer in 2002 and the second most common cause of cancer-related mortality (Parkin et al. 2005). There is considerable geographical variation in rates of gastric cancer, demonstrating higher rates in East Asia, Eastern Europe and parts of Central and South America. The lower risk areas are South Asia, North and East Africa, North America, Australia and New Zealand (Parkin et al. 2005).

The vast majority of stomach tumours are gastric adenocarcinomas (90%) which can be classified according to position (cardia or more distal tumours) and according to histological appearance (diffuse or intestinal). Tumours of the cardia are more likely to be diffuse type tumours with more distal tumours being intestinal type.

The bacterium Helicobacter pylori (H. pylori) is now known to be the predominant cause of chronic gastritis. Colonisation of the gastric epithelium by ‘unidentified curved bacilli’ and the association with chronic inflammation was first brought to the attention of the medical world by (Marshall 1983; Warren 1983) who also isolated the organism. Their achievement earned them the Nobel Prize in Physiology or Medicine in 2005. In addition to its role in chronic gastritis and peptic ulceration, H. pylori was recognised as a carcinogenic agent for gastric cancer by the International Agency for Research on Cancer (IARC 1994).

The bacterium itself is a spiral or slightly curved gram-negative rod, 2.4–4.0 μm in length, with 2–5 unipolar flagella which play a role in motility.

This microaerophilic organism colonises the gastric mucus gel and adheres to the epithelium. While generally thought to be non-invasive there is evidence that intra- and inter-cellular invasion occurs (Necchi et al. 2007). It has exceptional urease activity helping it survive in an acidic environment. The organism is mainly contracted in childhood predominantly by the gastric-oral route. The faecal–oral or the oral–oral route together with indirect transmission through contaminated food or water may be possible (Go 2002). A large study from the north of England showed that the most important factors for contracting H. pylori infection in the UK were worse socioeconomic conditions in childhood and number of siblings. The risk of infection rose with an increasing number of siblings (Moayyedi et al. 2002).

The prevalence of H. pylori infection in middle-aged adults was estimated as 74% in developing countries and 58% in developed countries (Parkin 2006). The same study suggested that 63.4% of gastric cancer cases worldwide were attributable to the bacterium, which translated to 5.5% of all cancers. H. pylori is most strongly associated with the non-cardia intestinal type adenocarcinoma and the rare gastric mucosa-associated lymphoid tissue-type (MALT) lymphoma (Wotherspoon et al. 1991). Other major risk factors for gastric cancer independent of H. pylori include pernicious anaemia, Epstein-Barr Virus (EBV) and previous gastric surgery. The following discussion relates to the complex interrelationship between H. pylori, inflammation and other associated risk factors in the pathogenesis of gastric adenocarcinoma.

2 Pathogenesis

Prior to the discovery of H. pylori it was well recognized that gastric inflammation and hypochlorhydria were associated with gastric cancer. Careful biopsy follow-up studies in high-risk populations had demonstrated the slow development of gastric atrophy over years and identified the importance of gastric body atrophy and intestinal metaplasia as risk factors for gastric cancer development (Siurala et al. 1966). Correa et al. (1975) developed the hypothesis that gastric cancer development was a slow and complex multistep process.

Fundamental to the model was the progression from normal gastric mucosa to non-atrophic gastritis, multifocal atrophic gastritis,and then the development of intestinal metaplasia and dysplasia before invasive cancer (Figs. 1, 2, 3, 4, 5, 6) (Correa et al. 1975).

Fig. 1
figure 1

Normal gastric mucosa (H & E 300× kindly supplied by Dr. Kevin West)

Full size image
Fig. 2
figure 2

H. pylori gastritis (H & E 300× kindly supplied by Dr. Kevin West)

Full size image
Fig. 3
figure 3

H. pylori and polymorpholeukocyte in gastric pit (modified Giemsa stain 600× kindly supplied by Dr. Judy Wyatt)

Full size image
Fig. 4
figure 4

Gastric intestinal metaplasia (H & E 300× kindly supplied by Dr. Kevin West)

Full size image
Fig. 5
figure 5

Gastric low-grade dysplasia (H & E 300× kindly supplied by Dr. Kevin West)

Full size image
Fig. 6
figure 6

Intestinal type gastric adenocarcinoma (H & E 300× kindly supplied by Dr. Kevin West)

Full size image

Helicobacter pylori gastritis is fundamentally the host immune response to the bacteria with neutrophils, lymphocytes and macrophages. Despite the pronounced host humoral and cell-mediated response, colonisation persists for decades. The immune response is thus ineffective, raising issues regarding immune evasion (reviewed by Wilson and Crabtree 2007). Although H. pylori is the cause of the initial gastric inflammation only a tiny proportion of infected individuals will ultimately progress to cancer. An important determinant of cancer risk relates to the phenotype of the H. pylori gastritis. Subjects with an antral predominant gastric inflammation maintain acid secretion and are not at risk of gastric cancer but are at risk of duodenal ulcer disease. Those individuals who develop an H. pylori pangastritis are at risk of developing multifocal atrophy, reduced acid secretion and an increased risk of gastric cancer. With decreasing acid secretion H. pylori colonisation may itself be lost.

Prior to the discovery of H. pylori, environmental factors were considered of utmost importance for gastric cancer pathogenesis. With the initial discovery of H. pylori the role of the bacterium took prominence. Recent studies have emphasised the importance of bacterial, host and environmental factors, and their complex interrelationship.

2.1 Bacterial Factors

There is considerable heterogeneity in terms of H. pylori strains. A number of putative virulence factors have been identified and recently reviewed (Amieva and El-Omar 2008). Currently most interest centres on membrane proteins related to adherence, the cytotoxin associated gene A (CagA) and the cytotoxin VacA.

CagA is a 120–130 kDa protein initially identified by studies investigating the gastric humoral response to H. pylori. Only a proportion of colonised subjects had an immune response to this protein and the presence of such a response was associated with the degree of inflammatory activity and mucosal damage (Crabtree et al. 1991). Many studies have subsequently demonstrated that subjects with CagA positive strains are more associated with peptic ulcer disease and gastric cancer. The CagA gene is part of a large pathogenicity island (cag PAI). Genes from this island encode the CagA protein and a type IV secretion system. Following bacterial adhesion the secretion system is responsible for the translocation of CagA bacterial protein into the host cell (Segal et al. 1999; Odenbreit et al. 2000). Once translocated into host cytoplasm, CagA has complex effects on the host cell both directly and following phosphorylation.

CagA phosphorylation occurs on tyrosine residuals at its five amino acid EPIYA (Glu-Pro-Ile-Tyr-Ala) repeat region by host cell kinases. The tyrosine phosphorylated CagA interacts with a number of host proteins which can activate the ERK/MAPK pathway (Keates et al. 1999; Ding et al. 2008). The effect is to alter cellular signalling, cell proliferation and differentiation, programmed death, cytoskeletal organisation, stress and inflammatory responses (Backert et al. 2001; Chang et al. 2006).

The N-terminus of unphosphorylated CagA can complex with several junction proteins (E-cadherin, ZO1 and JAM) resulting in disruption of the epithelial cell apical junction complex, loss of polarity and proinflammatory and mitogenic responses (Bagnoli et al. 2005; Amieva et al. 2003). This may include the development of intestinal metaplasia (Murata-Kamiya et al. 2007).

CagA is a complex protein associated with dramatic effects on epithelial cells and associated with an increased risk of gastric cancer, yet the majority of subjects colonised with CagA positive strains will never develop gastric cancer.

The H. pylori bacterial cytotoxin VacA was initially described by Leunk et al. (1988) who demonstrated that H. pylori broth supernatents caused vacuolisation of cultured cells. The VacA gene occurs in all strains but has considerable variability in terms of cytotoxin activity. The major variations occur in the VacA signal region (types s1 and s2), the mid region (types m1 and m2) and the more recently described intermediate region (i1 and i2) (Atherton et al. 1995; Rhead et al. 2007).

When the toxin is secreted, approximately 50% remains associated with the bacterial cell surface and gains access to the epithelium with bacterial adherence (Ilver et al. 2004). The VacA toxin has multiple toxigenic properties. By affecting endosome maturation and function it causes cytoplasmic vacuolisation and impairment of antigen presentation (Molinari et al. 1998). Other effects include apoptosis, inhibition of T cell activation, disruption of tight junctions and cytochrome c release (Gebert et al. 2003; Galmiche et al. 2000; Papini et al. 1998; Willhite and Blanke 2004). The s1 m1 strains are the most toxigenic and the most associated with gastric cancer (Basso et al. 2008). Information on the intermediate type is limited although the i1 type has been associated with cancer risk in an Iranian population (Rhead et al. 2007).

Adherence seems a critical step with regard to H. pylori virulence with the highest density of adherence around intercellular junctions (Steer 1984). Several outer membrane proteins appear to act as adhesins. One of the most studied is BabA which is a 78 kDa outer membrane protein which binds through Lewis b blood antigens on the gastric epithelium (Ilver et al. 1998). Subjects expressing BabA have a higher density of colonisation which is associated with an enhanced mucosal inflammatory response (Rad et al. 2002). Two genes are associated with the encoding of BabA, although only BabA2 is functionally active. The expression of BabA2 is associated with the s1 VacA subtype and when associated with CagA it is associated with an enhanced risk of cancer formation (Gerhard et al. 1999).

The bacterial virulence factors described above are byzantine and intriguing in terms of benefit to the bacteria and cancer risk for the host. Adherence would appear to be a necessary prerequisite for significant mucosal damage. The BabA2, CagA and s1/m1 VacA virulence factors appear to act synergistically in promoting inflammation and the development of intestinal metaplasia (Zambon et al. 2003). An attempt to quantify the relative risk of the CagA and VacA bacterial factors confirmed the previously described CagA and VacA polymorphisms with disease but found that the single most important factor was the number of EPIYA-C segments and hence the magnitude of CagA phosphorylation (Basso et al. 2008).

The risk factors described above are all associated with an increased risk of gastric cancer and detailed studies are unravelling the mechanisms of increased risk.

Subjects colonised with the highest risk bacterial genotype are still unlikely to develop gastric cancer, again emphasising the multifactorial nature of this disease.

2.2 Host Factors

Helicobacter pylori colonisation is responsible for the increased mucosal expression of many cytokines, including several interleukins (IL-1β, IL-6, IL-8), tumour necrosis factor-α (TNFα) and interferon-γ (IFN-γ) (Calam 1999). These have all been described as affecting the parietal cells of the stomach, inhibiting acid secretion. There are multiple polymorphisms in inflammation-associated genes, for example in the IL-1 gene cluster that results in an increased expression of IL-1β, and in the TNF-A gene, resulting in increased expression of TNF-α (Kusters et al. 2006). Polymorphisms present in the interferon-γ receptor 1 (IFNGR1) gene have been found to increase host susceptibility to H. pylori-associated gastric cancer development (Canedo et al. 2008), and it was postulated that an increased receptor presence would result in increased number of proinflammatory cytokines. IL-8 is an important cytokine responsible for activating and attracting neutrophils and lymphocytes and for the induction of proinflammatory cytokines. The IL-8-251 gene polymorphism is associated with increased IL-8 production in H. pylori-associated gastric mucosa with an increased risk of premalignant change (Li et al. 2010). The presence of three or four proinflammatory gene polymorphisms in association with H. pylori infection has been shown to increase the risk of gastric cancer substantially (El-Omar et al. 2003).

In addition to the cytokine polymorphisms described above, other polymorphisms have been associated with an increased risk of gastric cancer in association with H. pylori. These include polymorphisms of the Toll-like receptor TLR4, a type I transmembrane protein expressed in the gastric epithelium which acts as a lipopolysaccharide (LPS) receptor and activates proinflammatory pathways. Certain polymorphisms in TLR4 have been shown to increase the risk of premalignant and malignant change in H. pylori colonised individuals (Hold et al. 2007; Hishida et al. 2009). The gene selenoprotein S (SEPS1) encodes a protein in the endoplasmic reticulum which protects cells from oxidative damage and has a role in cytokine release (Curran et al. 2005).

A promoter polymorphism which results in reduced expression of SEPS1 has recently been associated with gastric cancer in a Japanese population (Shibata et al. 2009).

There is increasing interest in the potential role of bone marrow derived cells (BMDCs) as cancer stem cells in gastric neoplasia, reviewed by Correa and Houghton (2007). The bacterial and host factors resulting in ongoing inflammation and atrophy theoretically produce a situation where local stem cells are depleted. Mouse studies have demonstrated tissue stem cell depletion with ongoing inflammation and injury, and recruitment of bone marrow cells into the tissue stem cell niche. With continued inflammation and injury the BMDC-derived gastric mucosa developed metaplasia and dysplasia and subsequent invasive neoplasia (Houghton et al. 2004). The relevance and implications of these studies for human disease are as yet unclear.

2.3 Environmental Factors

Similar to other parts of the GI tract, the stomach has prolonged contact with ingested food material. The varying rates of gastric cancer geographically and chronologically have promoted an interest in the potential role of diet. Characteristically populations at high risk of gastric cancer have diets low in animal fats and proteins, low in fresh fruit and vegetables, high in starches and carbohydrates and high in salt and nitrates (Judd 1988). Particular interest has been paid to N-nitroso compounds as these have been found to induce tumours in the glandular stomachs of experimental animals and to the antioxidant vitamin C (ascorbic acid) as a potential protective factor.

Dietary N-nitroso compounds are typically found in cured meat, fish and beers, and have been associated with populations at an increased risk of gastric cancer (Judd 1988). There is also the potential for the formation of N-nitroso compounds in the stomach itself with gastric nitrate-reducing bacteria acting on dietary nitrates. H. pylori is not nitrate reducing, but if the H. pylori related inflammation involves the gastric body and results in decreased acid secretion, progressive atrophy can occur, raising the gastric pH. In these circumstances a range of other bacteria can colonise the stomach including nitrate-reducing bacteria and as a result higher levels of potentially carcinogenic N-nitroso compounds can be detected in gastric juice.

The recent EPIC-EUROGAST study demonstrated an association between endogenous formation of N-nitroso compounds and gastric cancer risk (Jakszyn et al. 2006). Paradoxically as the gastric pH rises, H. pylori colonisation is often lost, which could be due to a less favourable environment or competition from other bacteria. In practice H. pylori is absent from the stomach in most patients at the time when their cancer is diagnosed, emphasising the potential for varying factors to be important at different points in the natural history of gastric cancer.

Dietary ascorbic acid as an antioxidant has the potential to reduce carcinogenic N-nitrosamine formation in the stomach. Studies looking at gastric juice vitamin C levels demonstrated higher concentrations than would be expected from simple oral intake. These high levels are due to active gastric secretion of vitamin C. The mechanism for the secretion is unclear but like acid secretion it decreases with inflammation and atrophy (Rathbone et al. 1989; Sobala et al. 1989). Furthermore as the gastric luminal pH rises the proportion of the vitamin C available in its active form decreases thus increasing the risk of potentially carcinogenic N-nitroso compounds formation.

Another often overlooked environmental factor for gastric cancer is smoking with the EPIC study estimating that 17.6% of gastric cancer is related to smoking (González et al. 2003). A study looking at H. pylori status and smoking demonstrated a strongly increased risk of gastric cancer in those patients who were infected with CagA positive H. pylori and smoked, although the mechanism is unclear (Brenner et al. 2002).

3 Treatment

Both gastric cancer incidence and H. pylori infection rates are declining in many countries with the most probable reasons relating to improved living conditions. Important factors for this almost certainly include improved sanitation, water quality, food hygiene and a reduction in smoking. Despite the well-described association of gastric cancer risk with diets poor in vegetables and fruit, related intervention studies have failed to show a convincing benefit (Forman and Burley 2006).

A Cochrane analysis of randomised trials of anti-oxidant supplementation failed to demonstrate any evidence for a preventative effect against gastric cancer (Bjelakovic et al. 2008).

One of the sad facts regarding gastric cancer is the poor improvement in survival figures with treatment over the past 50 years. Clinical symptoms of gastric cancer generally occur late when the tumour is advanced and treatment options are limited. In the USA, Europe and China survival rates are only 20–25% whereas in Japan the equivalent survival rate is 52% (Parkin 2001). This difference is largely attributable to Japanese screening programmes to detect early disease. The current position regarding different treatment modalities has recently been well reviewed by Hartgrink et al. (2009).

With the important role H. pylori has in the early stages of the pathogenesis of gastric cancer it is compelling to believe that eradication programmes would be beneficial. The major problem here is that approximately 50% of the world’s population is infected with H. pylori, and only a tiny minority of these are at risk of gastric cancer. Certainly in low-risk populations such as the UK and most of Europe, broad-based screening and treatment programmes are unlikely to be cost effective. In high-risk populations such as Japan and Columbia the cost benefit ratios are much better and the Asia–Pacific consensus guidelines on gastric cancer prevention recommend screening and eradication in high-risk populations (Fock et al. 2008).

Current treatment regimens are not perfect with major problems regarding drug resistance which dramatically alters eradication rates. Where population treatment intervention studies have been carried out the results have been conflicting (Fuccio et al. 2007; Ito et al. 2009). For those with established atrophy and intestinal metaplasia, eradication is not so reliable in reducing cancer risk and endoscopic-histological screening is necessary to detect progression. This is demonstrated in a study on early gastric cancer patients treated by endoscopic mucosal resection and then treated with H. pylori eradication or placebo. At three years 3.3% of the eradicated patients had developed a metachronous gastric cancer compared to 8.8% of the placebo-treated patients (Fukase et al. 2008).

Clearly, advances in treatment regimens may alter the cost effectiveness of potential population treatment strategies but the development of an effective vaccine would be the best option for the future, but concerns remain regarding H. pylori’s immune evasion.

4 Conclusion

Gastric cancer remains a major health problem and is paradigmatic of a cancer which develops from inflamed mucosa. There is a clear pathway of development with many different factors operating. Despite a steady decline in the western world it remains a very significant cause of cancer deaths worldwide. Our understanding of the pathogenesis as described above has improved dramatically over the past thirty years however this has not translated into improved outcomes as yet.

The Correa pathway was an important concept in understanding the stepwise histological progression leading to invasive gastric cancer. Since this was described, the most fundamental breakthrough was the discovery of H. pylori and its role in causing histological chronic gastritis. Intensive research across the world has uncovered an astonishingly complex host–bacterial interaction and identified a number of interrelated factors pertinent to cancer risk (Fig. 7). The presence of H. pylori causes chronic mucosal inflammation early in the pathway. In the stepwise histological progression there appears to be a point of no return after which H. pylori plays no role and indeed colonisation is often lost. Eradication of or vaccination programmes against H. pylori in high-risk populations potentially offer a reduction in gastric adenocarcinoma and therefore the worldwide burden of cancer.

Fig. 7
figure 7

Modified Correa gastric cancer pathway

Full size image