Synopsis  

Gastritis is a diagnosis based on histological criteria; its aetiology is diverse. Helicobacter pylori colonization and non-steroidal anti-inflammatory drug consumption are among the most frequent and important. Most patients remain asymptomatic in the absence of complications (peptic ulcer disease). The major health implications associated with gastritis are attributable to longstanding disease, in particular gastric atrophy and related achlorhydria as well as gastric carcinogenesis. It remains a challenge to identify individuals in whom the disease will progress, enabling early intervention and disease prevention.

 Introduction   

Damage to the gastroduodenal mucosa, with and without the induction of circumscribed lesions, can be the result of bacterial infection, exogenous toxic factors, and/or autoimmune processes.

H. pylori infection and non-steroidal anti-inflammatory drugs (NSAIDS) are by far the two most prevalent aetiologies inducing gastroduodenal pathology. In the following, we discuss basic and clinical features related to these factors (Fig. 1).

 Gastritis   

Gastritis is a commonly described endoscopic finding, but accurate diagnosis and differentiation requires histological assessment. The currently accepted method of describing gastritis is the Sydney classification. Developed in 1996 and updated in 2001, it includes assessment of mucosal morphology, grade of atrophy, and the distribution of abnormalities (Fig. 2). Histological grading includes the activity of inflammation, grade of atrophy, presence of intestinal metaplasia, H. pylori infection and other particular aetiological agents. In addition the distribution of inflammatory changes can be defined as:

• antrum predominant;
  
• corpus predominant;
  
• pangastritis.
  

The phenotypes described should be considered as a gradual interconnecting spectrum rather than clear-cut distinct entities. Gastritis can be defined as either acute or chronic.

Acute gastritis may be induced by a variety of noxious agents, including alcohol, foods, medications, and infections. Acute gastritis often appears macroscopically as edema and reddening of the mucosa. Aspirin-induced forms are often described with characteristic punctate hemorrhages or fibrin-covered lesions in the antrum or duodenum. NSAID-induced lesions are often recognized as florid lesions in the prepyloric antrum. The histological findings include edema with capillary exudation of leucocytes (Figs 3–5).

Chronic gastritis may be classified into two major forms (1) H. pylori-associated and (2) H. pylori-unrelated forms. H. pylori-induced gastritis accounts for 90% of all forms of chronic gastritis (Fig. 6). Although typical endoscopic features have been described for individual forms of chronic gastritis, classification is frequently difficult by endoscopic appearance alone. In addition the often multifocal distribution of changes within the mucosa requires that multiple biopsies be obtained from different sites (Fig. 7). At least two biopsies should be taken from the antrum/pylorus region and two from the fundus/corpus region to enable accurate histological differentiation. Additional biopsies from the incisura are often not clinically relevant (although proposed in the last Sydney classification update). In addition, biopsies should be obtained from any visible lesions.

H.pylori-associated gastritis   

H. pylori infection is usually acquired in childhood by fecal–oral or oral–oral transmission. Humans represent the natural hosts for infection. The clinical manifestations usually occur later in life when patients present with symptoms (gastritis, peptic ulcer, or gastric cancer) and undergo investigation. H. pylori infection is associated with the development of a chronic active gastritis.

Common macroscopic appearances include spotty or streaky reddening of the mucosal folds in the antrum, with or without chronic erosions. Florid exudative erosions are uncommon in H. pylori infection. Following H. pylori colonization, the superficial gastric epithelium becomes infiltrated with neutrophils, lymphocytes, and plasma cells. With ongoing inflammation, mucosal epithelial regeneration occurs with mucus depletion and the development of focal intestinal metaplasia. Whether these latter abnormalities develop or not is dependent both on the virulence of the infecting strains and the host's genetic predisposition. This is clinically relevant as the type of underlying gastritis predisposes to the development of different disease entities (ulcers or carcinoma).

Gastritis: clinical manifestations and symptoms   

The majority of individuals (80%) with H. pylori infection and gastritis remain asymptomatic in the long term. It is not clear whether the presence of gastritis alone is sufficient to induce symptoms. The majority of patients will present with dyspepsia. Dyspepsia is a symptom complex referable to the upper gastrointestinal tract and includes epigastric pain, nausea, vomiting, and bloating. Despite several attempts at classifying dyspepsia there is still no clear way of predicting underlying endoscopic or histological findings based on any predominant symptom complex.

However, alarm symptoms may suggest the presence of serious underlying conditions (Fig. 8). It is important to note that gastritis refers to a histological entity and is not a symptom, although many patients and doctors use the term interchangeably. Many patients with gastritis will be asymptomatic and in addition some individuals may have symptoms without undergoing pathology. 'Gastritis' is an umbrella term for a variety of disease entities, and should be categorized based on histological and aetiological criteria.

 Ulcer disease   

H.pylori and ulcers   

H. pylori is the predominant pathogen in peptic ulcer disease. Supportive evidence comes from epidemiological data as well as from a change of pathogenetic events initiated by H. pylori in gastroduodenal mucosa and a complex puzzle of inflammatory events with repercussions in gastric physiology. The definitive proof for the causal role of H. pylori in peptic ulcer is that successful eradication results both in ulcer healing and prevention of disease relapse [1]. Epidemiological studies have attributed a fourfold increased risk of developing an ulcer for H. pylori-infected vs.-non-infected persons; this risk is estimated to be 25 times higher if the degree of inflammatory activity is high and antrum predominant [2,3]. Approximately 90% of patients with duodenal ulcer and 70% with gastric ulcers harbour H. pylori infection [4,5].

The difference in prevalence of H. pylori among duodenal and gastric ulcers is due to the higher frequency of NSAIDs, the other well-established aetiological agents frequently involved in gastric ulcers. The other side of the coin is that only about 10% of H. pylori-infected subjects in Western industrialized countries will eventually develop an ulcer; others develop symptoms without ulcers or, much more rarely, gastric neoplasia. Approximately 80% of H. pylori-infected persons will live with a persistent chronic gastric inflammation without symptoms and without any further progression to relevant clinical disease. The complexity of H. pylori as an ulcerogenic condition clearly suggests that additional factors are required for completion of the pathogenetic cascade towards ulcer formation.

H.pylori: the pathogenetic pathway   

The following factors are relevant in the pathogenetic cascade leading to peptic ulceration:

• pattern and phenotype of gastritis;
  
• alterations in the homeostasis of gastric hormones and acid secretion;
  
• gastric metaplasia in the duodenum as a prerequisite for H. pylori colonization;
  
• the interaction of H. pylori with the mucosal barrier and 'ulcerogenic' strains;
  
• therapeutic proof of causality;
  
• genetic factors.
  

Pattern and phenotype of gastritis in association with H. pylori   

Generally H. pylori colonizes the epithelium from the prepyloric antrum to the cardia, and the outcomes of H. pylori infection are significantly influenced by the topographical predominance of the chronic inflammation. The characteristic pattern of gastritis in patients with duodenal ulcer (DU) is the antrum-predominant distribution of H. pylori, with a high density of the bacteria and a high degree of inflammatory activity in this location [6–8]. Following eradication therapy, gastric mucosal alterations are usually fully reversible in DU [9–11].

In gastric ulcer (GU), the topographical expression of chronic gastritis is such that the body mucosa and antral mucosa are equally affected. Unlike DU, acid secretion is decreased in patients with GU, because of the more severe involvement of the body mucosa. Acid-inhibitory drugs, in particular protein pump inhibitors (PPIs), always lead to changes in the pattern of gastritis, with a reduction of bacteria and inflammatory activity in the antrum and an increase of both these conditions in the body.

Alterations in the homeostasis of gastric hormones and acid secretion related to H. pylori   

Antrum-predominant chronic H. pylori infection is accompanied by an increase in both basal and stimulated gastric acid output, the effect being most pronounced in patients with DU [12–14]. The link between H. pylori infection and acid hypersecretion appears to be hormonal in nature and driven by hypergastrinemia. The hypergastrinemia is a consequence of a reduction in gastric somatostatin synthesis and release, which normally exerts an inhibitory control on gastrin release [15,16]. More recent work has shown that, in addition to gastric hormones, neural pathways are also influenced by H. pylori infection, with functional disruption of antral–fundic neural connections [17]. The impairment of the inhibitory neural control of gastrin and acid secretion elicited by a test meal in patients with H. pylori-positive DU was recently demonstrated [18]. These abnormal responses, which in association with increased gastric emptying induce a higher duodenal acid load, are all reversed following H. pylori eradication. Eradication of H. pylori infection rapidly resolves the accompanying hypergastrinemia and is associated with normalization of antral somatostatin levels, whereas it may take several months before acid hypersecretion returns to normal [16,19,20]. Hypergastrinemia not only stimulates gastric acid secretion but also exerts a trophic effect on the parietal cell mass, which may explain the sustained acid secretion following H. pylori eradication.

Gastric metaplasia in the duodenum is a prerequisite for H. pylori colonization   

Gastric acid hypersecretion, and more specifically the acid overload in the duodenum, leads to the development of gastric metaplasia in the duodenal bulb. H. pylori is adapted to colonize gastric epithelial cells; as such, gastric metaplasia in the epithelium of the duodenal bulb is crucial and essential for the organisms to colonize the duodenum. Most likely this happens by migration from the antrum. The risk of developing a DU in the presence of gastric metaplasia and H. pylori appears to be 50 times that of controls [21].

Interaction of H. pylori with the mucosal barrier   

Colonization of the gastric mucosa by H. pylori evokes a local inflammatory response, which results in mucosal injury of varying degrees [22]. The acute inflammatory response starts with the release of epithelium-derived cytokines, predominantly interleukin-8 (IL-8) [23,24]. Bacterial products and cytokines orchestrate the acute inflammatory response, promoting an influx of neutrophils and macrophages into the gastric mucosa, which subsequently release lysosomal enzymes, leukotrienes, and oxygen-free radicals [23,25]. The next step in the inflammatory cascade is the activation of T and B lymphocytes by bacterial antigens with the release of further proinflammatory cytokines, including IL-1, IL-2, IL-6, and tumour necrosis factor-a (TNF-a), as well as the generation of specific antibodies (IgA and IgG) directed against H. pylori[26,27]. All these factors contribute to mucosal damage by promoting the release of factors with ulcerogenetic potential, such as platelet activating factor (PAF) and components of the complement pathway [28]

Ulcerogenic strains of H. pylori   

Pathogenetic properties (virulence factors) vary among strains. H. pylori isolated from patients with peptic ulcer disease appear to be more virulent, with some evidence that they exert a stronger adhesive property and produce greater amounts of enzymes with toxic potential. Among them, urease and phospholipases A2 and C are of particular interest. H. pylori strains from ulcer patients produce greater amounts of urease; this enzyme catalyzes the production of ammonia, which in high concentrations is followed by the formation of toxic complexes such as NH₄Cl [29]. Phospholipases A and C disrupt a defensive element of the mucosal barrier surfactant composed of a phospholipid-rich layer responsible for maintaining mucosal hydrophobicity, a key component in maintaining the integrity of the gastric epithelium [30,31]. The released phospholipids appear to be ingested by H. pylori[32].

Several other bacterial products are released in close proximity to the epithelial cells, exerting a direct harmful effect. Infection with certain H. pylori genotypes is linked to more severe morbidity. The most important genotypes are vacA- and cagA-positive, present in almost all patients with peptic ulceration. The H. pylori-derived vacuolating cytotoxin (vacA), an 87 kDa protein, causes vacuolar degeneration in cultured gastric cell preparations and gastric ulceration in experimental animals [33]. Although present in all H. pylori strains, the vacA gene, depending on its allelic form, is expressed in only 60%, lending strong support to the hypothesis of strain-dependent virulence. The cytotoxin-associated gene A (cagA), which is restricted to cytotoxin-producing strains of H. pylori, encodes for an 120–160 kDa immunodominant protein that is now recognized as a marker of greater virulence, leading to an enhanced local inflammatory response [34–37].

Among H. pylori-infected individuals, bacterial strains expressing cagA protein are closely associated with peptic ulcer (present in 92% of cases) and, to a lesser extent, with chronic gastritis (~60% in Western populations) [38,39]. The cagA gene is now known to be part of a larger cag pathogenicity island, which embraces other genes. These and a series of other genes (NAP, oiPA, iceA) have been characterized, and their function is to enhance mucosal inflammation through induction of cytokines [40,41].

Genetic factors and H. pylori   

Twin studies confirm a genetic predisposition for infection with increased susceptibility in monozygotes vs. dizygotes [42]. Lewis antigen association studies also support a genetic predisposition for peptic ulcer disease [43,44]. A Japanese working group contributed a new finding to this issue of genetic predisposition by reporting that subjects with human lymphocyte antigen (HLA) type DQA 1301 have an increased prevalence of ulcer disease [45].

The therapeutic proof of causality: H. pylori and ulcers   

The list of therapeutic studies is long, and all studies unequivocally report that H. pylori eradication leads to cure of gastric as well as duodenal ulcer [1,4,5]. The effect is lasting, as shown in long-term follow-up studies [46]. In addition, ulcer healing is accelerated if antibiotics are added to acid-secretory inhibitors [47]. Finally, complications of peptic ulcer disease can also be presented following H. pylori eradication, and this is superior to acid-inhibitory therapy alone (Fig. 9). All these facts taken together led to the NIH consensus statement in 1994 that antibiotic treatment in addition to antisecretory therapy is required for all patients with H. pylori-positive peptic ulcers [48].

Ulcers: clinical features and diagnosis   

Abdominal pain is the key presenting symptom in patients with either gastric or duodenal ulcers. A detailed description of the pain and its related features (intensity, site, radiation, relationship to meals, diurnal variation, aggravating and relieving factors, and associated symptoms such as nausea, vomiting, or weight loss) may assist differentiation between the types of peptic ulcer disease and also distinguish them from other gastric conditions such as gastric cancer.

It is particularly important to ascertain if there are any alarm symptoms, suggestive of sinister underlying pathology and indicating the need for rapid thorough investigation (Fig. 8). Typically, duodenal ulcers are characterized by pain at night, hunger pain, and pain which is relieved by eating. It is important however, to note that ulcers may be symptom free. Asymptomatic ulcers are more frequent in elderly subjects and those taking NSAIDs. Similarly, in patients on immunosuppressive therapy, such as corticosteroids, symptoms may be reduced or even masked completely.

Individuals with symptoms for four or more weeks require investigation. The use of radiological imaging techniques is now obsolete. The gold standard investigative tool is gastroscopy. Performing a gastroscopy facilitates accurate assessment of the lesion, including site, size, presence or absence of complications, and to distinguish between benign and malignant disease by performing biopsies (Figs 10–12).

Several successful management strategies have been proposed.

Test and treat   

The most accepted non-invasive management strategy is a so-called test and treat strategy[49]. In this strategy non-invasive H. pylori testing (C13 urea breath test, stool antigen assay, or serology) is undertaken as the first-line investigation in young people (<40 years) who present with dyspepsia (without alarm symptoms). Those who are H. pylori positive subsequently receive an accepted eradication therapy while those who are not infected receive empirical symptomatic treatment. Such a strategy has advantages in reducing the endoscopy workload but also in enabling rapid assessment and treatment of dyspeptic patients.

Endoscopic diagnosis   

Diagnosis of H. pylori at the time of gastroscopy requires both antral and corpus samples for histological assessment in combination with an additional antral sample for either rapid urease testing or culture. It is inadequate to perform only one diagnostic test. Again it should be stressed that histological samples from gastric ulcers should be obtained to detect malignancy. Following treatment (eradication therapy) a repeat gastroscopy is not indicated in the majority of patients, as a non-invasive H. pylori test performed at least 4 weeks after treatment has been shown to be effective. However, in the case of gastric ulcers, or complicated duodenal ulcers, a follow-up gastroscopy 6–8 weeks after instigating treatment is warranted both to ensure healing of the lesions and, in the case of gastric ulcers, to check again for malignancy.

 Treatment of peptic ulcers   

Acid suppression   

Acid suppression has been used for the management of patients with peptic lesions for over a hundred years. An historical breakthrough was the development of histamine 2 receptor antagonists (H2RAs), which had a greater effect on ulcer healing than any other modalities available up to then. The ability to suppress gastric acid secretion more profoundly became available with the development of omeprazole, a PPI, in 1989. PPIs dramatically enhanced ulcer healing rates: at 4 weeks 90% of ulcers could be healed compared to approximately 70% with H2RAs.

The scientific rationale behind employing acid suppression to induce ulcer healing is to attempt to maintain the gastric pH above 3.5, at which level pepsinogen enzymatic activation to pepsin, the injurious gastric factor, is inhibited. The need for long-term acid-suppressant therapy has been made redundant following the recognition that eradication of H. pylori infection results in a rapid and sustained healing of ulcer lesions with only a 5% risk of future relapse. H. pylori eradication therapy is now the principal treatment for both gastric and duodenal ulcers. A 7-day course of acid suppression combined with antibiotics is sufficient therapy in the case of uncomplicated ulcers. In those with a gastric ulcer or duodenal lesions >2 cm, signs of complications, or H. pylori-negative lesions, acid suppression therapy should continue until ulcer healing has been documented.

H.pylori eradication   

The currently recommended therapy for eradication is a short, 7 day combination of a PPI (standard dose bd) with clarithromycin and either amoxicillin or metronidazole, known as triple therapy [50–55]. Intention to treat analyses have confirmed that triple therapy induces ulcer healing in up to 80% of cases with successful eradication. The choice between combining clarithromycin (C) with either amoxicillin (CA) or metronidazole (CM) is difficult. Gastrointestinal side-effects, particularly diarrhoea, are more frequent when a CA combination is used (20 vs. 10%). CA would appear to be the treatment of choice in regions with high levels of primary metronidazole resistance (Fig. 13).

Treatment failure is frequently a result of either poor patient compliance (improved with the advent of shorter treatment duration) or antibiotic or PPI resistance. The issue of antimicrobial primary resistance has become increasingly important [56–58]. It has been estimated that the world-wide prevalence of primary resistance to metronidazole and clarithromycin is in the order of 20–70% and 1–12%, respectively. These figures vary dramatically by region. Amoxicillin resistance appears extremely uncommon. Secondary resistance to metronidazole has been demonstrated in 60–70% of individuals and to clarithromycin in 30–50%. Despite this, second-line treatments have been developed which offer eradication rates of 70–80% and which may be employed without necessitating culture and antibiotic sensitivity testing. If second-line therapy fails it is advisable to repeat a gastroscopy and to obtain mucosal samples for culture and sensitivity analysis, as retreatment is unlikely to be successful with empirical (blind] therapy (Fig. 14).

 NSAIDs and gastrointestinal pathology   

Introduction   

Gastrointestinal pathology associated with the use of NSAIDs is common and may be considered a public health issue. NSAID-related conditions are associated with significant morbidity and mortality. It has been estimated that over 30 million individuals regularly consume NSAIDs and that this figure is increasing rapidly. NSAIDs remain the mainstay of treatment for a myriad of inflammatory conditions and are frequently employed as analgesics; in particular the use of aspirin has risen dramatically since its efficacy in prevention of both coronary artery and cerebral vascular disease has been proven. The decreasing prevalence of H. pylori infection in the developed world coupled with the increased consumption of NSAIDs has meant that these drugs are now the most important cause of gastroduodenal ulceration in many countries.

Our understanding of the mechanisms involved in NSAID-induced gastrointestinal damage has expanded in recent times and led to the development of preventative strategies and novel therapeutic agents. The relative contributions of different pathophysiological processes described to date in the development of mucosal damage as a result of any particular NSAID is still a matter for discussion (Fig. 15). Similarly, whether data from endoscopic studies can be extrapolated to the wider population is not agreed. Although there is significant consensus in the literature regarding these issues, the presence of conflicting laboratory and clinical data has naturally resulted in some debate among the diverse groups of individuals who prescribe and manage complications of NSAIDs.

Clinical and histological characteristics of NSAID-related injury   

The consumption of NSAIDs induces acute gastric mucosal injury hours after ingestion, which is characterized by the presence of submucosal hemorrhage and the production of superficial mucosal erosions. Further ingestion may result in abrogation of these acute injuries via adaptation. Long-term consumption is associated classically with Type C gastritis, which is a relative misnomer as one of the key histological findings is a paucity of inflammatory response. The characteristic findings are epithelial, endothelial, and muscular hyperplasia.

Whether NSAID-induced gastritis is associated with symptoms such as dyspepsia is unclear. An observational study performed in England reported no increased prevalence of dyspepsia among NSAID users compared to controls over a 1 year follow-up. Conversely, a population-based study from America estimated that NSAID use was associated with a twofold increased risk of gastrointestinal symptoms. Differences in the definition of dyspepsia, and both the formulation and doses of NSAIDs employed, may account for these disparate results. A recent meta-analysis of 92 studies which reported gastrointestinal complications of NSAIDs attempted to address this issue. This showed that epigastric pain and related symptoms are increased in NSAID users by 36% compared to controls [59].

In many patients, ongoing tissue damage results in the development of deep excavated gastric or duodenal ulcers, partly as a direct result of mucosal injury but also due to impairment of mucosal healing mechanisms. The effects of NSAIDs are not confined to the gastric or duodenal mucosa. Consumption has been implicated in the development of esophagitis, esophageal strictures, acute intestinal ulceration, perforation, and, rarely, colonic ulceration. Often NSAID-related ulcers are large, more commonly found in the stomach than in the duodenum, and paradoxically frequently less painful than other peptic ulcers. It has been suggested that NSAID-associated peptic ulcers cause less pain because of their effect on the production of prostaglandins, which are known to contribute to the sensation of somatic and possibly visceral pain.

Epidemiology of NSAIDs and gastric injury   

It has been clearly demonstrated that patients using NSAIDs are at an increased risk of developing both ulcers and associated complications (hemorrhage or perforation). Consumption of NSAIDs is associated with a three- to fourfold increased risk of peptic ulcer development. The attributable risk of ulcer complications among NSAID users is 25–35%. Pooled data from several case control studies have reported that NSAID consumption is associated with relative risks of 3, 6, and 7.6 for gastrointestinal hemorrhage, perforation, and mortality, respectively [60].

Large-scale cohort studies have confirmed these findings [61–63]. In addition, there are now more data available on the gastrointestinal effects of prophylactic low-dose aspirin. Case control studies performed in the United Kingdom have shown that the use of low-dose aspirin confers an increased relative risk of bleeding compared to other patients and to the community of 2–6 depending on the dose employed (75–1200 mg/day) [64].

Further evidence from the UK Transient Ischaemic Attack Trial confirms previous data which showed that the risk of bleeding is greater for duodenal than gastric ulcers and that bleeding may occur from other sites of the gastrointestinal tract [65]. Moreover, the risk of duodenal perforation is also increased and has been reported to be six times that of non-users, but less than that for other NSAIDs [66]. It is worth noting that the concomitant use of low-dose aspirin and other NSAIDs has an additive effect and has been reported to be associated with twice the risk of perforation and bleeding.

Risk modifiers of injury with NSAIDs   

There are a number of possible risk modifiers that may affect the rates of peptic ulceration and their complications regionally (Fig. 16). The strength of available evidence for different risk modifiers varies. It has been clearly established that age, past history of peptic ulceration, high doses, and concomitant use of anticoagulants or corticosteroids magnify the risk of bleeding peptic ulceration.

Dosage and type of NSAID   

Patients with rheumatoid arthritis are more likely to develop complications of NSAIDs than individuals with osteoarthritis. Although some reports do suggest that disease entities confer differing risks it is more likely that the variation in risk reflects both the dosages and formulations regularly employed in different conditions. There are many reports available in the literature which demonstrate a linear association between the dose of NSAID and the development of ulcer complications [63,67,68]. A meta-analysis of 12 studies which investigated the effect of different NSAIDs on rates of ulcer complications demonstrated that different formulations did confer different risks (Fig. 17). However, despite there being a safety advantage for some formulations such as ibuprofen compared to piroxicam or azapropazone, this was not consistent at higher doses [69]. Other pharmacokinetic properties are also likely to alter the relative risk of varying formulations.

Age   

It has been clearly established that older age predisposes to peptic ulcer development. Studies have estimated the relative risk in subjects over 60 compared to controls to be in the order of 3.5. The effect of NSAIDs on this risk appears to be additive rather than synergistic. That is to say, both are associated with inherent risks but there does not appear to be any specific mechanism whereby the effect of one increases the deleterious effect of the other. The relative risk of ulcer disease in elderly individuals prescribed NSAIDs is approximately 13 [70].

Prior ulcer   

A previous history of peptic ulcer is a risk factor for subsequent ulcer development; the addition of NSAIDs appears to add to this risk, with estimates of a combined relative risk of 17. This effect is independent of H. pylori status and tends to be site specific. It is therefore plausible that abnormal healing as a result of disordered repair mechanisms may play a role in the interaction of these two risk factors.

Anticoagulants   

As with the aforementioned risk modifiers, the effect of combined usage of NSAIDs and anticoagulants reflects the additive risks inherent to both agents [71]. It is possible that different NSAIDs will interact to greater or lesser degrees. It would be logical to presume that aspirin would further enhance the effect because of its ability to interfere with platelet aggregation.

Corticosteroids   

The role of corticosteroids is not quite as clear-cut. Whether these agents are themselves ulcerogenic or merely enhance the effects of other noxious stimuli continues to be debated. There is evidence to show that regular usage of at least 10 mg of prednisone per day is associated, in combination with NSAIDs, with a threefold increased risk of ulcer bleeding compared to subjects taking NSAIDs alone [72].

H.pylori infection and NSAIDs combined   

There is conflicting evidence available in the medical literature regarding the nature of the relationship between H. pylori infection and NSAID-induced gastric toxicity. H. pylori infection is causally associated with the development of peptic ulcers, gastric cancer, and mucosal associated lymphoid tissue lymphoma. Contrary to the previously mentioned risk modifiers, there is evidence to show that concomitant infection may protect in some individuals and be deleterious in others using NSAIDs.

This apparent paradox can be explained partially by the diverse outcomes associated with infection itself. It is known that environmental, bacterial, and host factors combine to determine the long-term consequences of infection. Therefore similar factors are likely to interact in the relationship between H. pylori infection, NSAIDs, and the risk of disease development.

Evidence to support this theory comes from studies which have reported that H. pylori infection in patients at low risk of ulcer development and naive to NSAIDs appears to be at increased risk of ulceration compared to H. pylori-negative counterparts, whereas individuals with a past history of ulcers and on long-term NSAIDs appear to be protected by infection [73–75].

A plausible explanation for this apparent dichotomy is that H. pylori-produced cyclo-oxygenase (COX)-2 may promote the healing of NSAID-related ulcers, possibly by offsetting the systemic effects of NSAIDs on prostaglandin synthesis. Studies have also reported that the efficacy of PPIs is enhanced by the presence of H. pylori.

Whether to screen for and eradicate H. pylori in subjects taking long-term NSAIDs remains a controversial issue. However, there is growing consensus among medical practitioners that individuals with or without additional risk factors such as age and previous history of ulcer, who are likely to require treatment in the long term, should undergo testing for H. pylori, and eradication if positive [76,77]. Supporting evidence for such a strategy comes from studies which have reported an increased risk in such patients of early ulcer development and associated complications. Following successful eradication, patients with complicated NSAID-related ulceration or a past history will require maintenance therapy with other cytoprotectants. This may not be the case for patients taking low-dose aspirin, in whom recent evidence has shown that H. pylori eradication alone is sufficient to prevent ulcer relapse.

Management of NSAID-associated gastrointestinal toxicity   

Awareness of the possible risks inherent in the use of NSAIDs enables prescribing physicians to rationalize treatment according to simple basic principles. In particular the dose and type of NSAID employed should be considered carefully before initiating therapy. Increased doses are associated with an increased risk of complications, without necessarily improving their clinical efficacy.

Patient characteristics, such as past history of peptic ulcer disease (either related to NSAIDs or not), comorbidity (including maintenance therapies), age, and H. pylori status (if indicated according to previously stated criteria), should also be taken into consideration. In patients at an increased risk it is reasonable to try alternative treatments such as COX-2 inhibitors, or if necessary to use NSAIDs with primary pharmacological prophylaxis against complications.

Similarly, patients who have previously developed NSAID-related gastroduodenal complications should receive secondary prophylaxis while continuing to use these agents, even after successful eradication of Helicobacter infection.

The mainstay of treatment for active NSAID-related peptic ulcers remains profound acid suppression (Fig. 18).

Selective COX-2 inhibitors   

Recognition that COX occurred in two forms (a constitutive form, COX-1, responsible for many beneficial physiological functions, and an inducible form, COX-2, thought to mediate pathology) led to the production of selective COX-2 inhibitors with the aim of maintaining efficacy while reducing or eliminating adverse events.

The initial hypothesis has subsequently been shown to be anything but simple following further analysis. It is apparent from animal studies that COX-1 knockout mice do not develop spontaneous gastric pathology [78]. In addition, in patients with H. pylori infection, causally associated with the development of peptic ulcers, the predominant source of prostaglandin production is COX-1 despite COX-2 being an inducible form and increased in other forms of gastric inflammation. Conversely, COX-2 knockout mice have only minor changes in inflammatory processes and develop renal abnormalities [79].

It is also clear from additional research that inhibition of one isoform results in a corresponding increase in the other, suggesting that the two are not acting in isolation. Further supportive evidence comes from animal studies which have shown that inhibition of both COX-1 and -2 are required to induce ulceration [80].

Despite the fact that we are still a long way from fully understanding the functions of these isoforms and their interactions with each other and the gastric mucosa, there is good evidence to show that their use is associated with a reduced risk of both peptic ulcer development and associated complications. Even at supratherapeutic dosages, the available selective COX-2 inhibitors do not appear to induce acute mucosal injury or to interfere with gastric prostaglandin synthesis [81–83]. This is presumably the mechanism through which protection is conferred.

The therapeutic role of these various agents in inflammatory conditions continues to be evaluated. Their use has been associated with other reported NSAID-related side-effects such as salt and water retention, liver toxicity, and their efficacy and cost-benefit remain to be clarified. Furthermore there are data to show that these agents, by inhibiting COX-2, hinder the repair mechanisms involved in ulcer healing; although similar to other NSAIDs, this implies that they are not immune to gastrointestinal adverse reactions [84].

Similarly, whether administration of selective COX-2 inhibitors in patients with ulcerative colitis will result in disease exacerbation or relapse, as seen with other NSAIDs, is unknown. Although these issues remain to be clarified it is clear from the comparative studies undertaken to date that COX-2 inhibitors represent a significant advance in the area of gastrointestinal tolerability [85].

Prophylaxis against NSAID injury   

Available prophylactic agents include PPIs, the prostaglandin analogue misoprostol, and H2RAs. There is growing evidence to show that H2RAs are as effective as other available agents at preventing the development of duodenal ulcers, but less effective for gastric lesions [86]. For this reason it would appear that these agents would not be the preferred treatment modality. However, there are data to show that, at an increased dose, effectiveness is similar to that of omeprazole.

Misoprostol has been shown in a number of studies to reduce the risk of ulcer development in patients taking NSAIDs; the effect appears to be dose dependant. Unfortunately, significant side-effects are frequent in higher doses, in particular diarrhoea and abdominal pain. These effects appear to be reduced by the combination of misoprostol with diclofenac in a single tablet form, Arthrotec.

PPIs have been shown to reduce the occurrence of ulcers in patients on NSAIDs by 75–80% compared to placebo. In addition, comparative studies with H2RAs and misoprostol have revealed that omeprazole leads to a higher rate of ulcer healing and is also superior as a secondary prophylactic agent [87,88]. There is some evidence to suggest that misoprostol may be as effective as PPIs in a subgroup of H. pylori-negative patients. To date PPIs have the best established record for healing NSAID ulcers. Whether healing rates are quickest if the offending drug is stopped is not clear. NSAIDs are known to impair ulcer healing and it would therefore appear logical that improved quality and rates of healing would occur if they were discontinued.

 Conclusion   

H. pylori infection and NSAIDs are by far the commonest cause of gastroduodenal pathology.

H.pylori   

The discovery of H. pylori colonization of the gastric mucosa and its causal association with peptic ulcer disease has opened the way for a new and lasting treatment for these previously chronic and often complicated conditions. In addition the development of novel diagnostic tests and management strategies have revolutionized our approach to these diseases, enabling accurate diagnosis and subsequent treatment without recourse to invasive procedures, and often in the clinic environment.

NSAIDs   

NSAID-induced gastroduodenal toxicity is a common clinical problem associated with serious consequences. It remains the most frequently reported drug-related adverse event worldwide. Advances in recent years have led to a better understanding of the pathophysiological processes involved in mediating mucosal injury, and to the development of selective COX inhibitors which may protect patients from these sequelae. The precise role of these agents in a variety of clinical settings is still under investigation and studies examining their cost effectiveness are ongoing.

Prophylaxis   

Both primary and secondary prophylaxis trials have been reported, which have shown that, in high-risk groups, gastrointestinal complications can be reduced dramatically by a number of agents. H. pylori infection is not implicated in the development of NSAID-associated peptic ulceration, but data assessing its status as a risk modifier have reported contradictory results. Furthermore, although there is a growing consensus regarding the role of eradication therapy, both to prevent and manage these ulcers, it is far from unanimous, and will require additional investigation.

 Outstanding issues and future trends   

The discovery of H. pylori and its subsequent association with both peptic ulcer disease and gastric cancer in the 1980s profoundly altered the clinical management of these conditions. In particular surgical intervention for peptic ulcer disease has become a rare event, the majority of peptic ulcers being treated medically with eradication therapy. The development of powerful acid-suppressing medications facilitated the conservative treatment of peptic ulceration not induced by H. pylori, especially that related to NSAIDs.

More recently it has become medically accepted practice in geographical areas with a low prevalence of gastric cancer in young individuals to look for infection in people who have symptoms attributable to peptic ulcer disease by non-invasive means, thereby reducing the necessity for upper endoscopy, while providing appropriate treatment for infected individuals—the so-called 'test and treat strategy'. In addition to providing rapid, effective treatment for peptic ulcer disease and some patients with functional dyspepsia, there is a large body of evidence for this strategy having an excellent cost-benefit. This management strategy will, according to all available circumstantial evidence, reduce the incidence of long-term sequelae of infection, namely peptic ulcer and gastric cancer.

Development of search and treat management strategies for H. pylori infection would appear to be the most logical progression from current practice. A search and treat strategy can be retrospective, involving individuals with a history of peptic ulcer, and prospective, in that asymptomatic individuals at high risk of being infected and suffering complications would be screened for infection, especially those with a strong family history of distal gastric cancer. The widespread implementation of effective non-invasive diagnostic tests for H. pylori will facilitate such a development.

The future with regard to NSAID-related gastroduodenal pathology appears to be almost upon us. The development of new anti-inflammatory agents such as COX-2 inhibitors and their clinical application in a myriad of conditions should significantly reduce the incidence of serious gastrointestinal complications. It has also become apparent that novel antiplatelet agents offer a viable alternative for patients with ischemic heart disease and gastric side-effects from aspirin use, thereby alleviating a significant proportion of the burden of gastrointestinal disease associated with its widespread usage. Future challenges do exist. Firstly, the cost of these agents is substantial, and there is still need to optimize the strategies for the long-term use of PPIs as cytoprotectants. It is necessary to identify cost-benefit models for clearly defined patient groups.

 References  

1 Rauws, AEJ & Tytgat, GNJ. Cure of duodenal ulcer associated with eradication of Helicobacter pylori. Lancet 1990; 335: 1233–5. PubMed

2 Sipponen, P. Natural history of gastritis and its relationship to peptic ulcer disease. Digestion 1992; 51 (Suppl. 1): 70–5. PubMed

3 Schubert, TT, Bologna, SD & Nensey, Y et al. Ulcer risk factor: interactions between Helicobacter pylori infection, nonsteroidal use, and age. Am J Med 1993; 94: 413–18. PubMed

4 Malfertheiner, P & Bode, G. Helicobacter pylori and the pathogenesis of duodenal ulcer disease. Eur J Gastroenterol Hepatol 1993; 5: 51–8.

5 Tytgat, GNJ, Lee, A, Graham, DJ, Dixon, MF & Rokkas, T. The role of infectious agents in peptic ulcer disease. Gastroenterol Int 1993; 6: 76–89.

6 Neri, M, Susi, D, Bovani, I, Laerza, F, Mezzeti, A & Cuccurullo, F. Bacterial mucosal infiltration in Helicobacter pylori-associated gastritis: histological and clinical consequences. Am J Gastroenterol 1994; 89: 1801–5. PubMed

7 Khulusi, S, Mendall, MA, Patel, P, Levy, J & Northfield, TC. Is there a quantifiable difference in H. pylori infection density between DU and non-DU patients? Gastroenterology 1994; 106: A106.

8 Bayerdörffer, E, Miehlke, S, Lehn, N, Manes, GA, Sommer, A & Höchter, W et al. Chronic type B gastritis as an important denominator of peptic ulcer healing. Eur J Gastroenterol Hepatol 1993; 5 (Suppl. 3).

9 Graham, DY. Helicobacter pylori: consensus reached: peptic ulcer is on the way to becoming an historic disease. Am J Gastroenterol 1994; 89: 1137–9. PubMed

10 Stolte, M & Bethke, B. Elimination of H. pylori under treatment with omeprazole. Gastroenterology 1990; 28: 271.

11 Logan, RPH, Walker, MM & Misiewicz, JJ et al. Changes in the intragastric distribution of Helicobacter pylori during treatment with omeprazole. Gut 1995; 36: 12. PubMed

12 Levi, S, Beardshall, K, Haddad, G, Playford, R, Gosh, P & Calam, J. Campylobacter pylori and duodenal ulcers: the gastrin link. Lancet 1989; 1: 1167–8. PubMed

13 McColl, KEL & El-Omar, E. (1994). Effect of H. pylori infection on gastrin and gastric acid secretion. In: H. pylori: Basic Mechanisms to Clinical Cure (eds Hunt RH, Tytgat GNJ), p. 245. Kluwer, Dordrecht.

14 Moss, SF & Calam, J. Acid secretion and sensitivity to gastrin in patients with duodenal ulcers: effect of eradication of H. pylori. Gut 1993; 34: 888. PubMed

15 Moss, SF, Legon, S & Bishop, AE et al. Effect of H. pylori on gastric somatostatin in duodenal ulcer disease. Lancet 1992; 340: 930. PubMed

16 Odum, I, Petersen, HD, Andersen, IB, Hnasen, BF & Rehfeld, JF. Gastrin and somatostatin in H. pylori infected antral mucosa. Gut 1994; 35: 615–18. PubMed

17 Olbe, L, Hamlet, A, Dalenback, J & Fändrks, L. A mechanism by which H. pylori infection of the antrum contributes to the development of duodenal ulcer. Gastroenterology 1996; 110: 1386–94. PubMed

18 Hamlet, A & Olbe, L. The influence of H. pylori infection on postprandial duodenal acid load and duodenal bulb pH in humans. Gastroenterology 1996; 111: 391–400. PubMed

19 McColl, KEL, Fullarton, GM, Chittajallu, R, El Nujumi, AM, Macdonald, AMI, Dahill, SW & Hilditch, TE. Plasma gastrin, daytime intragastric pH and nocturnal acid output before and at 1 and 7 months after eradication of H. pylori in duodenal ulcer patients. Scand J Gastroenterol 1991; 26: 339–46. PubMed

20 El-Omar, E, Penman, I, Dorrian, CA, Ardill, JES & McColl, KEL. Eradicating H. pylori infection lowers gastrin mediated acid secretion by two thirds in patients with duodenal ulcer. Gut 1993; 34: 1060–5. PubMed

21 Bode, G, Malfertheiner, P, Mader, U, Stanescu, A & Ditschuneit, H. Fine structure of active and healed duodenal ulcer. Am J Gastroenterol 1991; 86: 179–86. PubMed

22 Blaser, MJ. Hypothesis on the pathogenesis and natural history of H. pylori-induced inflammation. Gastroenterology 1992; 102: 720–7. PubMed

23 Crabtree, JE, Peichl, P, Wyatt, JL, Stachl, U & Lindley, IJ. Gastric interleukin-8 and IgA IL-8 autoantibodies in H. pylori infection. Scand J Immunol 1993; 37: 65–70. PubMed

24 Crowe, SE, Alvarez, L, Dytoe, H, Hunt, RH, Müller, M & Sherman, P et al. Expression of interleukin 8 and CD54 by human gastric epithelium after H. pylori infection in vitro. Gastroenterology 1995; 108: 65–74. PubMed

25 Dunn, BE. Pathogenic mechanisms of Helicobacter pylori. Gastroenterol Clin North Am 1993; 22: 43–57. PubMed

26 Karttunen, R. Blood lymphocyte proliferation, cytokine secretion and appearance to T cells with activation surface markers in culture with H. pylori. Clin Exp Immunol 1991; 83: 396–400. PubMed

27 Hatz, R, Bayerdörffer, E, Lehn, N & Enders, G. Immune response in H. pylori infection: implications for treatment of gastroduodenal disease. Clin Immunother 1994; 2: 295–306.

28 Rosam, AC, Wallace, JL & Whittle, BJR. Potent ulcerogenic actions of platelet-activating factor on the stomach. Nature 1986; 319: 54–6. PubMed

29 Smoot, DT, Mobley, HLT, Chippendale, GR, Lewinson, JF & Resau, JH. H. pylori urease activity is toxic to human gastric epithelial cells. Infect Immun 1990; 58: 1992–4. PubMed

30 Goggin, PM, Marrero, JM, Spychal, RT, Jackson, PA, Corbishley, CM & Northfield, TC. Surface hydrophobicity of gastric mucosa in H. pylori infection: effect of clearance and eradication. Gastroenterology 1992; 103: 1486–90. PubMed

31 Mauch, F, Bode, G, Ditschuneit, H & Malfertheiner, P. Demonstration of a phospholipid-rich zone in the human gastric epithelium damaged by Helicobacter pylori. Gastroenterology 1993; 105: 1698–704. PubMed

32 Hills, BA. Gastric mucosal barrier: evidence for H. pylori in gastric surfactant and deriving protection from it. Gut 1993; 38: 2071–80.

33 Figura, N, Guglielmetti, A, Rossolino, A, Barberi, A, Cusi, G & Musmanno, RA et al. Cytotoxin production by Campylobacter pylori strains isolated from patients with peptic ulcers and from patients with chronic gastritis only. J Clin Microbiol 1989; 27: 225–6. PubMed

34 Covacci, A, Censini, S, Bugnoli, M, Petracca, R, Burroni, D & Maccina, G et al. Molecular characterisation of the 128-kDa immunodominant antigen of H. pylori associated with cytotoxicity and duodenal ulcer. Proc Natl Acad Sci USA 1993; 90: 5791–5. PubMed

35 Cover, TL & Blaser, MJ. Purification and characterization of the vacuolating toxin from H. pylori. J Biol Chem 1992; 267: 10570–5. PubMed

36 Tummuru, MK, Sharma, SA & Blaser, MJ. H. pylori cagC, a homolog of the Bordetella pertussis toxin secretion protein, is required for induction of IL-8 in gastric epithelial cells. Gastroenterology. : A246.

37 Peek, RM, Miller, GG, Tham, KT, Perez-Perez, GI, Zhao, XM & Atherton, JC et al. Heightened cytokine expression and inflammatory response in vivo to cagA H. pylori strains. Lab Invest 1995; 73: 760–70. PubMed

38 Atherton, JC. The clinical relevance of strain types of H. pylori. Gut 1997; 40: 701–3. PubMed

39 Blaser, MJ. Intrastrain differences in H. pylori a key question in mucosal damage. Ann Med 1995; 27: 559–63. PubMed

40 Censini, S, Lange, C, Xiang, Z, Crabtree, JE, Ghiara, P, Borodovsky, M, Rappuoli, R & Covacci, A. Cag A pathogenicity island of H. pylori, encodes type I-specific and disease-associated virulence factors. Proc Natl Acad Sci USA 1996; 93: 14648–53. PubMed

41 Van Doorn, L-J, Figueiredo, C, Sanna, R, Plaisier, A, Schneeberger, P, De Boer, W & Quint, W. Clinical relevance of the cagA, vacA and iceA status of H. pylori. Gastroenterology 1998; 115: 58–66. PubMed

42 Malaty, HM, Engstrand, M, Pedersen, NL & Graham, DY. H. pylori infection: genetic and environmental influences. Ann Intern Med 1994; 120: 982–6. PubMed

43 Cowan, WK. Genetics of duodenal and gastric ulcer. Clin Gastroenterol 1973; 2: 539–46.

44 Boren, T, Falk, P, Roth, KA, Larson, G & Normark, S. Attachment of H. pylori to human gastric epithelium mediated by blood group antigens. Science 1993; 262: 1892–5. PubMed

45 Azuma, T, Konishi, J, Tanaka, Y, Hirai, M, Ito, S, Kato, T & Kohli, Y. Contribution of HLA-DQA gene to host's response against H. pylori. Lancet 1994; 343: 542–3. PubMed

46 Forbes, GM, Glaser, ME, Cullen, DJE, Warren, JR, Christiansen, KJ & Marshall, BJ et al. Duodenal ulcer treated with H. pylori eradication: seven year follow-up. Lancet 1994; 343: 258–60. PubMed

47 Huang, JQ, Wilkinson, JM & Hunt, RH. Antimicrobials accelerate healing of duodenal ulcer (DU) when given with proton pump inhibitors (PPI) or H2-receptor antagonists (H2RA): a meta-analysis. Gut, 1997; 41: A93.

48 NIH Consensus Conference. H. pylori in peptic ulcer disease. NIH Consensus Development Panel on H. pylori Peptic Ulcer Disease. JAMA 1994; 272 (1): 65–9. PubMed

49 Malfertheiner, P, Megraud, F, O'Morain, C, Hungin, AP, Jones, R, Axon, A, Graham, DY & Tytgat, GNJ and the H. pylori Study Group (EHPSG). Current concepts in the management of H. pylori infection: the Maastricht 2000 consensus report. Aliment Pharmacol Ther 2002; 16: 167–80. PubMed

50 Bazzoli, F, Zagari, RM & Fossi, S et al. Short-term low-dose triple therapy for the eradication of H. pylori. Eur J Gastroenterol Hepatol 1994; 6: 773–7.

51 Jaup, BH & Norrby, A. Low-dose, short term triple therapy for cure of H. pylori infection and healing of peptic ulcers. Am J Gastroenterol 1995; 90: 943–5. PubMed

52 Labenz, J, Stolte, M & Rühl, GH et al. One week, low-dose triple therapy for the cure of H. pylori infection. Eur J Gastroenterol Hepatol 1994; 7: 9–11.

53 Moayyedi, P, Sahary, P & Tompkins, DS et al. Efficacy and optimum dose of omeprazole in a new 1-week triple therapy regimen to eradicate H. pylori. Eur J Gastroenterol Hepatol 1995; 7: 835–40. PubMed

54 Adamek, RJ, Szymanski, C & Pfaffenbach, B et al. Short term triple therapy with pantoprazole, clarithromycin and metronidazole for the healing of H. pylori infection. Dtsch Med Wochenschr 1995; 120: 358–60. PubMed

55 Misiewicz, JJ, Harris, AW, Bardhan, KD, Levi, S, O'Morain, C & Cooper, BT et al. One week triple therapy for H. pylori: a multicentre comparative study. Gut 1997; 41: 735–9. PubMed

56 Lind, T, Veldhuyzen van Zanten, S & Unge, P et al. Eradication of H. pylori using one-week triple therapies combining omeprazole with two antimicrobials: the MACH I study. Helicobacter 1996; 1: 138–44. PubMed

57 Megraud, F. (1994). H. pylori resistance to antibiotics. In: H. pylori: Basic Mechanisms to Clinical Cure (eds Hunt RH, Tytgat GNJ), pp. 570–83. Kluwer, Dordrecht.

58 Lind, T, Megraud, F, Unge, P, Bayerdörffer, E, O'Morain, C & Spiller, R et al. The MACH 2 study: the role of omeprazole in eradication of H. pylori with One-Week Triple Therapies: a randomised double-blind study. Submitted.

59 Straus, WL, Ofman, JJ, MacLean, C, Morton, Berger, ML, Roth, EA & Shekelle, P et al. Do NSAIDs cause dyspepsia? A meta-analysis evaluating alternative dyspepsia definitions. Am J Gastroenterol 2002; 97: 1951–8. PubMed

60 Hawkey, CJ. Non-steroidal anti-inflammatory drugs and peptic ulcers. BMJ 1990; 300: 278–84. PubMed

61 Fries, J, Williams, CA, Bloch, DA & Michel, BA. Non-steroidal anti-inflammatory drug-associated gastropathy: incidence and risk factor model. Am J Med 1991; 91: 213–22. PubMed

62 Griffin, MR, Piper, JM, Daugherty, JR, Snowden, M & Ray, WA. Non-steroidal anti-inflammatory drug use and increased risk for peptic ulcer disease in elderly persons. Ann Intern Med 1991; 114: 257–63. PubMed

63 Wolfe, MM, Lichtenstein, DR & Sing, G. Gastrointestinal toxicity of non-steroidal anti-inflammatory. N Engl J Med 1999; 340: 1888–99. PubMed

64 Weil, J, Colin-Jones, D, Langman, M, Lawson, D, Logan, R & Murphy, M et al. Prophylactic aspirin and risk of peptic ulcer bleeding. BMJ 1995; 310: 827–30. PubMed

65 Slattery, J, Warlow, CP, Shoeeock, CJ & Langman, MJS. Risks of gastrointestinal bleeding during secondary prevention of vascular events with aspirin: analysis of gastrointestinal bleeding during the UK-TIA Trial. Gut 1995; 37: 509–11. PubMed

66 Lanas, A, Serrano, P, Bajador, E, Esteva, F, Benito, R & Sainz, R. Evidence of aspirin use in both upper and lower gastrointestinal perforation. Gastroenterology 1997; 112: 683–9. PubMed

67 Gabriel, SE, Jaakkimainen, L & Bombardier, C. Risk for serious gastrointestinal complications related to the use of nonsteroidal anti-inflammatory drugs: a meta-analysis. Ann Intern Med 1991; 115: 787–96. PubMed

68 Langman, MJS, Weil, J, Wainwright, P, Lawson, DH, Rawlins, MD & Logan, RF et al. Risk of bleeding peptic ulcer associated with individual non-steroidal anti-inflammatory drugs. Lancet 1994; 343: 1075–8. PubMed

69 Henry, D, Lim, LL-Y, Garcia Rodriquez, LA, Gutthann, SP, Carson, JL & Griffin, N et al. The variability in risk of gastrointestinal complications with individual non-steroidal anti-inflammatory drugs: results of a collaborative meta-analysis. BMJ 1996; 312: 1563–6. PubMed

70 Garcia Rodriguez, LA & Jick, H. Risk of upper gastrointestinal bleeding and perforation associated with individual non-steroidal anti-inflammatory drugs. Lancet 1994; 343: 769–72. PubMed

71 Shorr, RI, Ray, WA, Daugherty, JR & Griffin, MR. Concurrent use of non-steroidal anti-inflammatory and oral anticoagulants places elderly persons at high risk for haemorrhagic peptic ulcer disease. Arch Intern Med 1993; 153: 1665–70. PubMed

72 Piper, JM, Ray, WA, Daugherty, JR & Griffin, MR. Corticosteroid use and peptic ulcer disease: role of nonsteroidal anti-inflammatory drugs. Ann Intern Med 1991; 114: 735–40. PubMed

73 Hawkey, CJ, Tulassay, Z, Szczepanski, L, van Rensburg, CJ, Filipanciz-Susnowska, A & Lanas, A et al. Randomised controlled trial of Helicobacter pylori eradication in patients on non-steroidal anti-inflammatory drugs. HELP NSAID's Study: Helicobacter Eradication for Lesion Prevention Lancet 1998; 352: 1016–21. PubMed

74 Chan, FK, To, KF, Wu, JC, Yung, MY, Leung, WK & Kwok, T et al. Eradication of Helicobacter pylori and risk of peptic ulcers in patients starting long term treatment with non-steroidal anti-inflammatory drugs: a randomised trial. Lancet 2002; 359: 9–13. PubMed

75 Labenz, J, Blum, AL, Bolten, WW, Dragosics, B, Rosch, W, Stolte, M & Koelz, HR. Primary prevention of Diclofenac associated ulcers and dyspepsia by Omeprazole or triple therapy in Helicobacter positive patients: a randomised double blind, placebo controlled trial. Gut 2002; 51: 329–35. PubMed

76 Sung, JJY, Russell, RI, Yeomans, N, Chan, FKL, Chen, SL & Fock, KM et al. Working Party Report, Non-steroidal anti-inflammatory drug toxicity in the upper gastrointestinal tract. J Gastroenterol Hepatol 2000; 15: G58–68. PubMed

77 Langenbach, R, Morham, SG, Tiano, HF, Loftin, CD, Ghanayem, BI & Chulada, PC. Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell 1995; 83: 483–92. PubMed

78 Dinchuk, JE, Car, BD, Focht, RJ, Johnston, JJ, Jaffee, BD & Covington, MB et al. Renal abnormalities and an altered inflammatory response in mice lacking cyclooxygenase II. Nature 1995; 378: 406–9. PubMed

79 Tanaka, A, Araki, H, Hase, S, Komoike, Y & Takeuchi, K. Upregulation of COX II by inhibition of COX I in the rat: a key to NSAID induced gastric injury. Aliment Pharmacol Ther 2002; 16: 90–101. PubMed

80 Lanz, FL, Rack, MF, Simon, TJ, Quan, H, Bolognese, JA & Hoover, ME et al. Specific inhibition of cyclooxygenase-2 with MK-0966 is associated with less gastroduodenal damage than either aspirin or ibuprofen. Aliment Pharmacol Ther 1999; 13: 761–7. PubMed

81 Laine, L, Harper, S, Simon, T, Bath, R, Johanson, J & Schwartz, H et al. A randomised trial comparing the effect of rofecoxib, a cyclooxygenase-2 specific inhibitor, with that of ibuprofen on the gastroduodenal mucosa of patients with osteoarthritis. Gastroenterology 1999; 117: 776–80. PubMed

82 Wight, NJ, Garlick, N, Calder, N, Dallob, A, Gottesdiener, K & Hawkey, CJ. Evidence that the COX-2 specific inhibitor rofecoxib at 50mg spares gastric mucosal prostaglandin synthesis in humans. Gut Supplement 1999; V(45): 30–3.

83 Futagami, S, Hiratsuka, T, Wada, K, Tatsuguchi, A, Tsukui, T & Miyake, K et al. Inhibition of Helicobacter pylori induced cyclo oxygenase-2 aggravates NSAID- caused gastric damage in Mongolian Gerbils. Aliment Pharmacol Ther 2002; 16: 847–55. PubMed

84 Stover, RR. Are selective COX 2 inhibitors superior to traditional NSAIDs? Both CLASS and VIGOR trials support the COX 2 hypothesis. BMJ 2002; 324: 1287–8. PubMed

85 Koch, M, Capurso, L, Dezi, A, Ferrario, F & Scapingnato, C. Prevention of NSAID-induced gastroduodenal mucosal injury: meta-analysis of clinical trials with Misoprostol and H2-receptor antagonists. Dig Dis Sci 1995; 1: 62–74.

86 Yeomans, ND, Tulassay, Z, Juhasuz, L, Racz, I, Howard, JM & van Rensburg, CJ et al. for the ASTRONAUT Study Group. A comparison of Omeprazole and ranitidine for treating and preventing ulcers associated with non-steroidal anti-inflammatory drugs. N Engl J Med 1998; 338: 719–26. PubMed

87 Agrawal, N, Safdi, M, Wruble, L, Karvois, D, Greski-Rose, P & Huang, B. Effectiveness of lansoprazole in the healing of NSAID induced gastric ulcer in patients continuing to take NSAIDs. Gastroenterology 1998; 114: A52.

88 Hawkey, CJ, Karrasch, JA, Szcepanski, L, Walker, DG, Barkun, A, Swannell, AJ & Yeomans, ND, for the omeprazole Vs Misoprostol for NSAID-Induced Ulcer Management (OMNIUM) Study Group. Omeprazole compared with Misoprostol for ulcers associated with non-steroidal anti-inflammatory drugs. N Engl J Med 1998; 338: 727–34. PubMed

89 Graham, DY, Heppo, KS, Remirez, FC, Lew, GH & Saeed, ZA. Treatment of Helicobacter pylori reduces the rate of rebleeding peptic ulcer disease. Scand J Gastroenterol 1993; 28: 939–42. PubMed

90 Labenz, J & Borsch, G. Role of Helicobacter pylori eradication in the prevention of peptic ulcer bleeding relapse. Digestion 1994; 55: 19–23. PubMed

91 Jaspersen, D, Koerner, T, Schorn, W, Bremensthal, M, Reischka, C & Hammar, CH. Helicobacter pylori eradication reduces the rate of rebleeding in ulcer haemorrhage. Gastrointest Endosc 1995; 41: 5–7. PubMed

92 Reimann, JF, Schilling, D, Schauwecker, P, Wehlen, G, Dorlars, D, Kohler, B & Meier, M. Cure with Omeprazole plus Amoxicillin versus long-term Ranitidine therapy in Helicobacter pylori associated peptic ulcer bleeding. Gastrointest Endosc 1997; 46: 371–3. PubMed

93 Rokkas, T, Karameris, A, Mavrogeorgis, A, Rallis, E & Giannikos, N. Eradication of Helicobacter pylori reduces the possibility of rebleeding in peptic ulcer disease. Gastrointest Endosc 1995; 41: 1–4. PubMed

94 Santander, C, Gravales, RG, Gomez-Cedenilla, A, Cantero, J & Pajares, JM. Antimicrobial therapy for Helicobacter pylori infection versus long-term maintenance antisecretion treatment in the prevention of recurrent haemorrhage from peptic ulcer: prospective nonrandomised trial on 125 patients. Am J Gastroenterol 1996; 91: 1549–52. PubMed

Copyright © Blackwell Publishing, 2004