Ulcers and gastritis

• Introduction
• Realistic expectations from gastroprotective therapy for NSAID users
• Protective effect of antisecretory drugs and nitrates on ulcer bleeding
• Effect of PPIs on re-bleeding risk among high-risk users taking COX-2 inhibitors
• COX-2 inhibitors and cardiovascular complications
• Gastritis and gastric cancer
• New frontiers in the eradication of H. pylori
• Key learning points
• References

Introduction

Upper gastrointestinal bleeding (UGIB) remains a major problem worldwide. Over time the management issues have changed and a paradigm shift occurred when it became clearly established that eradication of Helicobacter pylori would ”cure” peptic ulcers, including the elimination of new ulcers, recurrent ulcers, and ulcer complications. The subsequent recognition that H. pylori infections increased the risk of complications from nonsteroidal anti-inflammatory drugs (NSAIDs) led to the recommendation to test NSAID and aspirin users for H. pylori, and to eradicate the infection whenever it is found. Unfortunately, eradication of H. pylori is no longer straightforward, as clinicians are facing the increasing challenge of resistant strains. Moreover, eradication of H. pylori alone will not solve the problem of NSAID ulcer complications because NSAIDs themselves increase the risk of UGIB. Unresolved issues include identification of the circumstances in which particular traditional NSAIDs or cyclo-oxygenase-2 (COX-2)-selective inhibitors are preferred, and the actual role for gastroprotective co-therapy.

Although there are many studies on gastroprotective therapies, their clinical effectiveness remains unclear. This is due, in part, because (with few exceptions) pharmaceutical company-sponsored trials have used surrogate markers instead of clinically relevant outcome trials in high-risk patients. When one searches for data, one finds that pharmaceutical company-sponsored consensus statements appear to outnumber relevant clinical trials. The available data have also focused on relative instead of absolute risk or risk reduction, and are therefore of limited use for clinical decision making. Clinical decision making has been further complicated by the recent realization that the benefits obtained from NSAID use may be offset by an increase in cardiovascular complications (e. g. myocardial infarctions). This review highlights new data that address these current issues. Five recent papers have been selected for more detailed analyses [1] [2] [3] [4] [5].

Realistic expectations from gastroprotective therapy for NSAID users

Current data regarding the potential benefits of proton pump inhibitors (PPIs) in the prevention of re-bleeding in H. pylori-negative high-risk patients, who continue to be NSAID users, have yielded disappointing results, with re-bleeding occurring in approximately 10 % [6] [7] [8]. Overall the data suggest that the actual risk reduction in re-bleeding is in the range of 50 % to 75 % ([Fig. 1]). There are now epidemiological studies which show that this appears to be a reasonable estimate.

Protective effect of antisecretory drugs and nitrates on ulcer bleeding

Lanas et al. performed a retrospective case-control study that examined the effect of antisecretory drugs and nitrates on the risk of ulcer bleeding associated with NSAIDs, antiplatelet agents, and anticoagulants [1]. Data were collected on the use of nonselective NSAIDs, COX-2 inhibitors, aspirin, non-aspirin antiplatelet agents, and anticoagulants among 2777 patients admitted with endoscopically confirmed peptic ulcer bleeding. PPIs, H₂-receptor antagonists (H2RAs), and transdermal or oral nitrates were associated with a 67 %, 35 %, and 48 % reduction, respectively, in the overall risk for developing UGIB. NSAID or aspirin use was associated with an increased relative risk (RR) of UGIB of 5.6 (95 % confidence interval [CI] 5.0-6.3); the use of PPIs, H2RAs, and nitrates in this subset led to an 82 %, 61 %, and 49 % reduction in UGIB risk, respectively. Compared with H2RAs, PPI use was more effective in reducing UGIB risk in patients taking traditional NSAIDs, aspirin, or clopidogrel. Importantly, none of the antisecretory agents had a significant effect in reducing UGIB risk in patients taking anticoagulants.

The study has a few limitations. The number of patients taking concomitant PPIs, H2RAs, or nitrates was low in some subsets. In addition, 81 % of the study cases were estimated to have H. pylori infection, likely providing an overestimation of the risk reduction provided by antisecretory agents, compared with what is likely to be achieved in populations where most are H. pylori negative. No conclusions on the long-term benefits and risks can be drawn, as analyses were restricted to drugs of interest taken in the week prior to the UGIB event. The study also did not address co-therapy (e. g. aspirin and clopidogrel), which is especially important considering the combined use of aspirin and clopidogrel in patients with coronary stents.

The estimated degree of benefit of prophylactic PPIs among NSAID users was also similar to another recent study reporting a 60 % reduction in hospitalization due to peptic ulcer among diclofenac users (odds ratio [OR] 0.4; 95 % CI 0.2 - 0.7; P < 0.05) [9]. Overall, the available data are consistent with the notion that prophylactic PPI use among NSAID users will reduce the risk of UGIB by 50 % - 80 %, and are in line with the results of the prospective clinical trials from Hong Kong [6] [7] [8].

Effect of PPIs on re-bleeding risk among high-risk users taking COX-2 inhibitors

COX-2 inhibitors have been branded as safer in terms of gastrointestinal toxicity than traditional NSAIDs. How much safer are they? One study suggested the relative risk of upper gastrointestinal complications in COX-2 inhibitor use was 2.6 (95 % CI: 1.9 - 3.6) compared with 3.7 (95 % CI: 3.1 - 4.3) in traditional NSAID users [10]. However, it is generally agreed that the potential gastrointestinal benefits of COX-2 inhibitors are offset by concomitant aspirin use. Chan et al. conducted a single-center, double-blind, randomized placebo-controlled trial studying the effect of adding a prophylactic PPI to celecoxib, in 273 individuals at high risk for UGIB [2]. Entry criteria included endoscopically confirmed NSAID-induced UGIB, confirmation of ulcer healing, and H. pylori eradication where appropriate. Patients were randomized to receive either high-dose celecoxib (200 mg twice daily) or in combination with esomeprazole (20 mg twice daily) for 12 months. The 15 % of patients in each arm at risk for cardiovascular events were allowed low-dose (80 mg) aspirin for primary prevention. After a median follow-up of 13 months (range 0.4 - 13 months), significantly more patients assigned to celecoxib only experienced recurrent gastroduodenal ulcer bleeds (i. e. 8.9 % vs. 0 %), compared with patients assigned to celecoxib and esomeprazole (P = 0.0004). A total of 83 % of ulcers recurred at their previous locations. The addition of aspirin in the control group increased the re-bleeding rate from 7.1 % to 19 % over 13 months, in line with previous reports suggesting that concomitant aspirin with COX-2 inhibitors may negate its apparently superior gastrotoxicity profile [10] [11]. Recurrent bleeding was seen in none of those receiving celecoxib, a PPI, and aspirin (P = 0.03). However as only 15 % of patients in each arm took concomitant aspirin, larger studies will be needed to determine the effect of PPI on aspirin and COX-2 inhibitor co-therapy.

This study confirmed prior results showing a high rate of recurrent bleeding among high-risk patients receiving celecoxib [6]. The 0 % (95 % CI 0 - 206) rate of recurrent UGIB is unprecedented and suggests that high-risk individuals who need ongoing NSAID therapy can best be treated by a combination of celecoxib and a PPI. This combination appears not to be associated with the unacceptably high rate of re-bleeding among high-risk patients receiving traditional NSAIDs and PPIs.

Most authorities recommend that the decision regarding the use of gastroprotective agents be based on risk stratification. Unfortunately, tables showing absolute risk are generally unavailable and those that are available have not been tested prospectively. In the average-risk patient, PPI prophylaxis is not recommended, in part because the risk for UGIB in these patients is very low and drug costs can be enormous. However, gastroprotective agents are recommended for higher-risk patients (e. g. age > 65 years). For example, if one achieved a 75 % reduction in bleeding in a high-risk group (e. g. 10 % bleeding per year), and the cost of the PPI was $ 1, the cost for PPI alone would be only about $ 5000 per bleeding event prevented (number needed to treat [NNT] = 13). The drug costs would range up to about $ 38 000 for a double dose of a high-cost PPI and would therefore still be in the ”acceptable” range. Most high-risk groups have about 5 % bleeding risk per year and thus the costs would double. For true lower-risk patients (e. g. age < 65 years) the NNT and therefore the costs would increase 10- to 20-fold and be unacceptably high, hence the recommendation to restrict co-therapy to high-risk groups.

The major costs associated with a NSAID-induced UGIB are those associated with hospitalization, which are constant for any given risk reduction. Thus, the major difference in cost in prevention strategies is the cost of the gastroprotective therapy. The available endoscopic studies have consistently shown a lack of a dose-response effect in terms of prevention. If the data are surrogates for clinical outcomes, the lowest dose of the least expensive PPI should be chosen. For very high-risk patients (prior NSAID-associated bleed, prior ulcer bleed, prior clinical NSAID ulcer), the risk of re-bleeding is so high (e. g. > 10 % would be expected to re-bleed and the mortality rate would be at least 1 %), that avoidance of NSAIDs would clearly be the preferred strategy over using an NSAID with PPI prophylaxis. However, as shown above, the combination of celecoxib plus a PPI may provide a reasonable alternative strategy. It is important to note that in the epidemiological study of Lanas et al. [1], PPIs failed to provide protection in patients taking anticoagulants. Therefore the benefits of PPI and celecoxib co-therapy may be reduced in patients taking concomitant dicumarinics; these patients may still be at increased risk of UGIB.

COX-2 inhibitors and cardiovascular complications

The Vioxx Gastrointestinal Outcomes Research (VIGOR) study [12] reported a significant increase in the incidence of myocardial infarction in patients taking rofecoxib compared with patients taking naproxen, and raised concerns regarding the adverse cardiovascular events associated with COX-2 inhibitors. Risk estimates regarding the cardiovascular hazards of NSAIDs, their relation to COX-1/COX-2 prothrombotic properties and its impact on blood pressure, congestive cardiac failure, and glomerular filtration rate are now becoming available. Abraham et al. provided data from a large retrospective cohort study of 384 322 veterans who were NSAID or COX-2-selective NSAID users [3]. NSAIDs were stratified according to their degree of COX-2 selectivity; the incidences of myocardial infarction and cerebrovascular events (CVAs) were calculated for the ensuing 3 months. Rofecoxib and valdecoxib were classified as highly COX-2-selective agents. Current use of highly selective NSAIDs was associated with a 47 % increase (adjusted) in myocardial infarction compared with low-selectivity agents. There was also a significant association with highly selective agents for CVAs (adjusted hazards ratio [HR] 1.6; 95 % CI 1.2 - 2.2) with the majority being ischemic in nature. The increased risk of highly selective NSAIDs was seen in both low and average cardiovascular risk patients. Rofecoxib conferred the greatest risk for myocardial infarction and CVA in both risk groups. There was no difference in the risk of myocardial infarction and CVAs between patients on moderately selective NSAIDs and those on poorly selective NSAIDs. Patients with no active exposure to NSAIDs had the least risk of myocardial infarction and CVAs.

This study has two implications. Firstly, the risk of cardiovascular events increases with any active exposure to an NSAID, even after adjustment for existing cardiovascular risk factors. Secondly, this increase in risk is proportional to the COX-2 selectivity of the NSAID, contradicting previous theories on this being a class effect. These results are consistent with another large study that demonstrated an elevated risk of cardiovascular events in rofecoxib users compared with a reduced rate in naproxen users, irrespective of the baseline cardiovascular risk [13]. Whether the latter drug actually provides a protective effect remains unclear, and data defining relative and absolute cardiovascular and gastrointestinal risks of individual NSAIDs are eagerly awaited. Although this study did not find a significant association with other COX-2 inhibitors and NSAIDs, many complex mechanisms are likely at play. It is also thought that interindividual variability in the response to COX-2 inhibitors, possibly due to polymorphisms in COX and CYP2CP, may play a part in an individual’s susceptibility to the cardiovascular complications conferred by COX-2 inhibitors [14].

Gastritis and gastric cancer

H. pylori is now accepted as the major cause of gastric cancer [15]. The risk is related to the extent and severity of gastric mucosal atrophy. The Operative Link on Gastritis Assessment (OLGA) group proposed a standardized staging system to estimate the risk of gastric malignancy. The staging system is based on the topography and severity of gastritis and was recently validated in 439 patients [16]. The prevalence of H. pylori infection (active or eradicated) increased progressively from Stage 0 to Stage IV (51 % vs. 100 %; P = 0.003). Benign conditions (including duodenal ulcers) consistently clustered in Stage 0 - II, whereas all neoplastic lesions clustered in stages III - IV. The system, though appearing simple, is open to interobserver variability and requires targeted biopsy mapping. The costs involved in screening all patients with Stages III - IV gastritis are likely to be substantial. Studies are needed to address whether it has advantages over simpler methods, such as using changes in pepsinogen levels to assess the severity and extent of atrophy and thus cancer risk [17].

Many studies and surveillance programs use age as the surrogate for risk. Take et al. prospectively followed 1131 patients after H. pylori eradication [4]. Annual endoscopies were done to study the relationship between baseline gastritis and the subsequent risk of gastric cancer. Baseline gastric mucosal atrophy was evaluated endoscopically according to the endoscopic-atrophic border scale described by Kimura and Takemoto [18], and classified into mild, moderate, and severe grades, depending on the topographic location of the visualized atrophy. Patients with persistent H. pylori infection had a significantly higher risk of 70/100 000 per year of developing gastric cancer compared with a risk of 23/100 000 per year among those patients cured of H. pylori (P = 0.04). All 13 gastric cancers developed in patients with preceding stomach ulcers. The risk of developing gastric cancer increased with the severity of the baseline gastric atrophy (P = 0.01). Persistent H. pylori infection (HR 3.9; P = 0.03), baseline gastric mucosal atrophy (HR 3.3; P = 0.01), and age (HR 2.0; P = 0.04) were significant factors for the risk of developing gastric cancer. This study confirmed that the risk of gastric cancer appears to be proportional to the extent and degree of mucosal atrophy. Importantly, one can conclude that risk can be predicted by the extent and degree of mucosal atrophy, whether assessed endoscopically, by histology, OLGA staging, or by pepsinogen testing, and that age alone should likely be abandoned as a criterion for gastric cancer screening programs. Overall the available data support the notion that H. pylori eradication before significant expansion of atrophy takes place will largely prevent gastric cancer; eradication after development of significant atrophy will only reduce the rate of cancer development. Among the choices for population screening (e. g. pepsinogen testing, OLGA staging, or endoscopic-atrophic border detection), only pepsinogen testing appears suitable and cost-effective.

New frontiers in the eradication of H. pylori

The success rates of H. pylori eradication treatment have dropped considerably over recent years largely due to the increasing emergence of antibiotic resistance. For instance, clarithromycin resistance is particularly common in Western countries (up to 25 %). Legacy triple therapy using a PPI, amoxicillin, and clarithromycin rarely yields 80 % eradication rates (intention to treat). Sequential therapy, consisting of 5-day dual therapy (PPI plus amoxicillin) followed by 5-day triple therapy (PPI, clarithromycin, and tinidazole) has repeatedly been shown to be superior, but whether the mechanism is simply related to the addition of the fourth drug or some other mechanism remains unclear. Vaira et al. recently confirmed that sequential therapy was more effective in patients with clarithromycin-resistant strains than legacy triple therapy (89 % vs. 29 %; P = 0.0034) but failed in the face of dual clarithromycin and metronidazole resistance [5]. Prior studies evaluating all four drugs (PPI, amoxicillin, clarithromycin, metronidazole) administered concomitantly had shown similar high cure rates consistent with the notion that the benefit was related to the addition of the fourth drug [19] [20]. Overall, one must conclude that in countries where clarithromycin resistance is common, legacy triple therapy should no longer be used unless pretreatment susceptibility testing is available. Sequential therapy is preferred. The addition of tinidazole/metronidazole to legacy triple therapy (triple-based quadruple therapy) is a theoretical option but before it can be widely recommended, additional trials are needed. Sequential therapy is not ideal and studies of dose, duration, and number and type of drug are clearly needed. Excellent results continue to be reported with high-dose (e. g. q. i. d.) PPI and amoxicillin dual therapy, providing more aggressive acid inhibition, which has been shown to influence H. pylori eradication rates [21].

Key learning points

• Among gastroprotective agents, PPIs appear most efficacious in reducing UGIB risk but the degree of protection is only in the range of a 50 % to 80 % reduction. This protection is not apparent in patients taking anticoagulants.

• Celecoxib and PPI co-therapy could be considered as an alternative in high-risk patients who need ongoing NSAID therapy.

• NSAID use increases the risk of cardiovascular events. NSAIDs appear to differ with regard to risk, with naproxen appearing to be safest, and the very highly COX-2-selective NSAIDs conferring the highest risk.

• The risk of gastric cancer is proportional to the degree and severity of gastric atrophy. Eradication of H. pylori reduces cancer risk; eradication before significant expansion of atrophy produces a much greater reduction in cancer risk than does eradication after atrophy is established. The degree of atrophy and not age should be the key criterion for gastric cancer screening programs.

• Sequential therapy is more effective than legacy triple therapy for eradication of clarithromycin-resistant H. pylori strains. Sequential therapy can clearly be improved.

Competing interests: None. Dr. Graham is a member of the executive committee of the PRECISION study that is evaluating the cardiovascular risks of celecoxib, naproxen and ibuprofen in high cardiovascular risk patients. Dr. Graham receives no compensation for this activity.

References

• 1 Lanas A, Garcia-Rodriguez L A, Arroyo M T. et al . Effect of antisecretory drugs and nitrates on the risk of ulcer bleeding associated with nonsteroidal anti-inflammatory drugs, antiplatelet agents, and anticoagulants.  Am J Gastroenterol. 2007;  102 507-515
  CrossrefPubMedSearch in Google Scholar
• 2 Chan F K, Wong V W, Suen B Y. et al . Combination of a cyclo-oxygenase-2 inhibitor and a proton-pump inhibitor for prevention of recurrent ulcer bleeding in patients at very high risk: a double-blind, randomized trial.  Lancet. 2007;  369 1621-1626
  CrossrefPubMedSearch in Google Scholar
• 3 Abraham N S, El-Serag H B, Hartman C. et al . Cyclooxygenase-2 selectivity of non-steroidal anti-inflammatory drugs and the risk of myocardial infarction and cerebrovascular accident.  Aliment Pharmacol Ther. 2007;  25 913-924
  PubMedSearch in Google Scholar
• 4 Take S, Mizuno M, Ishiki K. et al . Baseline gastric mucosal atrophy is a risk factor associated with the development of gastric cancer after Helicobacter pylori eradication therapy in patients with peptic ulcer diseases.  J Gastroenterol. 2007;  42 21-27
  CrossrefPubMedSearch in Google Scholar
• 5 Vaira D, Zullo A, Vakil N. et al . Sequential therapy versus standard triple-drug therapy for Helicobacter pylori eradication: a randomized trial.  Ann Intern Med. 2007;  146 556-563
  PubMedSearch in Google Scholar
• 6 Chan F K, Hung L C, Suen B Y. et al . Celecoxib versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis.  N Engl J Med. 2002;  347 2104-2110
  CrossrefPubMedSearch in Google Scholar
• 7 Chan F K, Chung S C, Suen B Y. et al . Preventing recurrent upper gastrointestinal bleeding in patients with Helicobacter pylori infection who are taking low-dose aspirin or naproxen.  N Engl J Med. 2001;  344 967-973
  CrossrefPubMedSearch in Google Scholar
• 8 Lai K C, Chu K M, Hui W M. et al . Celecoxib compared with lansoprazole and naproxen to prevent gastrointestinal ulcer complications.  Am J Med. 2005;  118 1271-1278
  CrossrefPubMedSearch in Google Scholar
• 9 Hoer A, Gothe H, Schiffhorst G. et al . Comparison of the effects of diclofenac or other non-steroidal anti-inflammatory drugs (NSAIDs) and diclofenac or other NSAIDs in combination with proton pump inhibitors (PPI) on hospitalization due to peptic ulcer disease.  Pharmacoepidemiol Drug Saf. 2007;  16 854-858
  CrossrefPubMedSearch in Google Scholar
• 10 Garcia-Rodriguez L A, Barreales Tolosa L. Risk of upper gastrointestinal complications among users of traditional NSAIDs and COXIBs in the general population.  Gastroenterology. 2007;  132 498-506
  CrossrefPubMedSearch in Google Scholar
• 11 Lanas A, Garcia-Rodriguez L A, Arroyo M T. et al . Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-oxygenase-2 inhibitors, traditional non-aspirin non-steroidal anti-inflammatory drugs, aspirin and combinations.  Gut. 2006;  55 1731-1738
  CrossrefPubMedSearch in Google Scholar
• 12 Bombardier C, Laine L, Reicin A. et al . Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.  N Engl J Med. 2000;  343 1520-1528
  CrossrefPubMedSearch in Google Scholar
• 13 Solomon D H, Avorn J, Sturmer T. et al . Cardiovascular outcomes in new users of coxibs and nonsteroidal antiinflammatory drugs: high-risk subgroups and time course of risk.  Arthritis Rheum. 2006;  54 1378-1389
  CrossrefPubMedSearch in Google Scholar
• 14 Fries S, Grosser T, Price T S. et al . Marked interindividual variability in the response to selective inhibitors of cyclooxygenase-2.  Gastroenterology. 2006;  130 55-64
  CrossrefPubMedSearch in Google Scholar
• 15 Uemura N, Okamoto S, Yamamoto S. et al . Helicobacter pylori infection and the development of gastric cancer.  N Engl J Med. 2001;  345 784-789
  CrossrefPubMedSearch in Google Scholar
• 16 Rugge M, Meggio A, Pennelli G. et al . Gastritis staging in clinical practice: the OLGA staging system.  Gut. 2007;  56 631-636
  CrossrefPubMedSearch in Google Scholar
• 17 Watabe H, Mitsushima T, Yamaji Y. et al . Predicting the development of gastric cancer from combining Helicobacter pylori antibodies and serum pepsinogen status: a prospective endoscopic cohort study.  Gut. 2005;  54 764-768
  CrossrefPubMedSearch in Google Scholar
• 18 Kimura K, Takemoto T. An endoscopic recognition of the atrophic border and its significance in chronic gastritis.  Endoscopy. 1969;  3 87-97
  Thieme ConnectPubMedSearch in Google Scholar
• 19 Nagahara A, Miwa H, Ogawa K. et al . Addition of metronidazole to rabeprazole-amoxicillin-clarithromycin regimen for Helicobacter pylori infection provides an excellent cure rate with five-day therapy.  Helicobacter. 2000;  5 88-93
  CrossrefPubMedSearch in Google Scholar
• 20 Treiber G, Ammon S, Schneider E. et al . Amoxicillin / metronidazole / omeprazole / clarithromycin: a new, short quadruple therapy for Helicobacter pylori eradication.  Helicobacter. 1998;  3 54-58
  CrossrefPubMedSearch in Google Scholar
• 21 Sugimoto M, Furuta T, Shirai N. et al . Evidence that the degree and duration of acid suppression are related to Helicobacter pylori eradication by triple therapy.  Helicobacter. 2007;  12 317-323
  CrossrefPubMedSearch in Google Scholar

D. Y. Graham, MD 

Michael E. DeBakey Veterans Affairs Medical Center

RM 3A-320 (111D)2002 Holcombe Boulevard
Houston
TX 77 030
USA

Fax: +1-713-790-1040

Email: dgraham@bcm.tmc.edu

References

• 1 Lanas A, Garcia-Rodriguez L A, Arroyo M T. et al . Effect of antisecretory drugs and nitrates on the risk of ulcer bleeding associated with nonsteroidal anti-inflammatory drugs, antiplatelet agents, and anticoagulants.  Am J Gastroenterol. 2007;  102 507-515
  CrossrefPubMedSearch in Google Scholar
• 2 Chan F K, Wong V W, Suen B Y. et al . Combination of a cyclo-oxygenase-2 inhibitor and a proton-pump inhibitor for prevention of recurrent ulcer bleeding in patients at very high risk: a double-blind, randomized trial.  Lancet. 2007;  369 1621-1626
  CrossrefPubMedSearch in Google Scholar
• 3 Abraham N S, El-Serag H B, Hartman C. et al . Cyclooxygenase-2 selectivity of non-steroidal anti-inflammatory drugs and the risk of myocardial infarction and cerebrovascular accident.  Aliment Pharmacol Ther. 2007;  25 913-924
  PubMedSearch in Google Scholar
• 4 Take S, Mizuno M, Ishiki K. et al . Baseline gastric mucosal atrophy is a risk factor associated with the development of gastric cancer after Helicobacter pylori eradication therapy in patients with peptic ulcer diseases.  J Gastroenterol. 2007;  42 21-27
  CrossrefPubMedSearch in Google Scholar
• 5 Vaira D, Zullo A, Vakil N. et al . Sequential therapy versus standard triple-drug therapy for Helicobacter pylori eradication: a randomized trial.  Ann Intern Med. 2007;  146 556-563
  PubMedSearch in Google Scholar
• 6 Chan F K, Hung L C, Suen B Y. et al . Celecoxib versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis.  N Engl J Med. 2002;  347 2104-2110
  CrossrefPubMedSearch in Google Scholar
• 7 Chan F K, Chung S C, Suen B Y. et al . Preventing recurrent upper gastrointestinal bleeding in patients with Helicobacter pylori infection who are taking low-dose aspirin or naproxen.  N Engl J Med. 2001;  344 967-973
  CrossrefPubMedSearch in Google Scholar
• 8 Lai K C, Chu K M, Hui W M. et al . Celecoxib compared with lansoprazole and naproxen to prevent gastrointestinal ulcer complications.  Am J Med. 2005;  118 1271-1278
  CrossrefPubMedSearch in Google Scholar
• 9 Hoer A, Gothe H, Schiffhorst G. et al . Comparison of the effects of diclofenac or other non-steroidal anti-inflammatory drugs (NSAIDs) and diclofenac or other NSAIDs in combination with proton pump inhibitors (PPI) on hospitalization due to peptic ulcer disease.  Pharmacoepidemiol Drug Saf. 2007;  16 854-858
  CrossrefPubMedSearch in Google Scholar
• 10 Garcia-Rodriguez L A, Barreales Tolosa L. Risk of upper gastrointestinal complications among users of traditional NSAIDs and COXIBs in the general population.  Gastroenterology. 2007;  132 498-506
  CrossrefPubMedSearch in Google Scholar
• 11 Lanas A, Garcia-Rodriguez L A, Arroyo M T. et al . Risk of upper gastrointestinal ulcer bleeding associated with selective cyclo-oxygenase-2 inhibitors, traditional non-aspirin non-steroidal anti-inflammatory drugs, aspirin and combinations.  Gut. 2006;  55 1731-1738
  CrossrefPubMedSearch in Google Scholar
• 12 Bombardier C, Laine L, Reicin A. et al . Comparison of upper gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis.  N Engl J Med. 2000;  343 1520-1528
  CrossrefPubMedSearch in Google Scholar
• 13 Solomon D H, Avorn J, Sturmer T. et al . Cardiovascular outcomes in new users of coxibs and nonsteroidal antiinflammatory drugs: high-risk subgroups and time course of risk.  Arthritis Rheum. 2006;  54 1378-1389
  CrossrefPubMedSearch in Google Scholar
• 14 Fries S, Grosser T, Price T S. et al . Marked interindividual variability in the response to selective inhibitors of cyclooxygenase-2.  Gastroenterology. 2006;  130 55-64
  CrossrefPubMedSearch in Google Scholar
• 15 Uemura N, Okamoto S, Yamamoto S. et al . Helicobacter pylori infection and the development of gastric cancer.  N Engl J Med. 2001;  345 784-789
  CrossrefPubMedSearch in Google Scholar
• 16 Rugge M, Meggio A, Pennelli G. et al . Gastritis staging in clinical practice: the OLGA staging system.  Gut. 2007;  56 631-636
  CrossrefPubMedSearch in Google Scholar
• 17 Watabe H, Mitsushima T, Yamaji Y. et al . Predicting the development of gastric cancer from combining Helicobacter pylori antibodies and serum pepsinogen status: a prospective endoscopic cohort study.  Gut. 2005;  54 764-768
  CrossrefPubMedSearch in Google Scholar
• 18 Kimura K, Takemoto T. An endoscopic recognition of the atrophic border and its significance in chronic gastritis.  Endoscopy. 1969;  3 87-97
  Thieme ConnectPubMedSearch in Google Scholar
• 19 Nagahara A, Miwa H, Ogawa K. et al . Addition of metronidazole to rabeprazole-amoxicillin-clarithromycin regimen for Helicobacter pylori infection provides an excellent cure rate with five-day therapy.  Helicobacter. 2000;  5 88-93
  CrossrefPubMedSearch in Google Scholar
• 20 Treiber G, Ammon S, Schneider E. et al . Amoxicillin / metronidazole / omeprazole / clarithromycin: a new, short quadruple therapy for Helicobacter pylori eradication.  Helicobacter. 1998;  3 54-58
  CrossrefPubMedSearch in Google Scholar
• 21 Sugimoto M, Furuta T, Shirai N. et al . Evidence that the degree and duration of acid suppression are related to Helicobacter pylori eradication by triple therapy.  Helicobacter. 2007;  12 317-323
  CrossrefPubMedSearch in Google Scholar

D. Y. Graham, MD 

Michael E. DeBakey Veterans Affairs Medical Center

RM 3A-320 (111D)2002 Holcombe Boulevard
Houston
TX 77 030
USA

Fax: +1-713-790-1040

Email: dgraham@bcm.tmc.edu