Proton-Pump Inhibitors: Clinical Applications

Introduction

The use of proton-pump inhibitors (PPIs) for the treatment of gastrointestinal acid-secretory disorders such as gastroesophageal reflux disease (GERD) and peptic ulcer disease (PUD) has dramatically altered the management of these conditions. PPIs are widely acknowledged to be the most effective acid-suppressing medications available and are considered first-line therapy for the symptoms of GERD and the maintenance of esophageal healing in patients with erosive esophagitis (EE). Currently, there are 5 PPIs approved by the US Food and Drug Administration (FDA) for use in this country. All of the PPIs are administered orally, with the exception of pantoprazole, which is available in both oral and intravenous (IV) formulations (ie, the only PPI approved for use in the United States as an IV formulation). Recently, 1 PPI, omeprazole, was approved for over-the-counter use and is now available without a prescription for patients with intermittent GERD symptoms. The continuing clinical experience with the entire class of PPIs, as well as refinements in our understanding of acid-peptic disorders, has led to new questions about the roles of PPIs in clinical practice.

During this year's meeting of the American College of Gastroenterology (ACG), several investigators reviewed the clinical indications and outcomes associated with the use of both IV and oral PPIs. This article reports on these presentations, with a view toward their implications for clinical practice.

Pharmacology of PPIs

Physiologic control of acid secretion in the stomach is controlled by a number of redundant second messenger pathways activated as a result of the binding of gastrin, acetylcholine, histamine, and prostaglandins on the basolateral surface of parietal cells. The ultimate factor in acid secretion, however, is stimulation of the proton pump (H⁺, K⁺ -ATPase) to secrete hydrogen ions into the gastric lumen in exchange for potassium ions. The substituted benzimidazoles (PPIs) all work to inhibit H⁺, K⁺ -ATPase in the same manner. As a weak base, the benzimidazole accumulates in the secretory canaliculus of the acid-secreting parietal cell, where it is rapidly protonated to the active form. This active form of the PPI binds to a sulfhydryl group on the H⁺, K⁺ -ATPase (proton pump) and prevents secretion of acid into the gastric lumen.

All of the PPIs are effective for the symptomatic control of GERD and healing of EE. Multiple studies have been completed and are currently under way to determine whether there are clinically important differences between individual PPIs. Over the last several years there has been an increase in the use of these agents in the critical-care setting, primarily for stress ulcer prophylaxis and complicated PUD, and significant research efforts have also been directed toward these novel indications.

Clinical Applications of IV PPIs

IV PPIs in Patients With Bleeding PUD*

Gastrointestinal hemorrhage from PUD is responsible for more than 150,000 hospitalizations and 8000-12,000 deaths per year. Much of the morbidity and mortality that accrues from bleeding PUD arises from rebleeding. The risk of rebleeding can be estimated from the endoscopic appearance of bleeding peptic ulcer lesions. This risk is approximately 4% with a clean-based ulcer, 33% with an adherent clot over an ulcer, 50% with a visible vessel, and 80% to 90% with an oozing or actively bleeding ulcer.^[1] Endoscopic therapy of these lesions has proven to be effective in reducing the risk of rebleeding, urgent surgery, and mortality.^[2] The addition of acid suppression with IV PPIs has also been shown to further decrease the risk of rebleeding in patients requiring endoscopic therapy in several well-designed trials.^[3,4] Patients enrolled in these trials had high-risk lesions, including active bleeding, visible vessel, or adherent clot. Patients with low-risk lesions, such as clean-based ulcers, are at such low risk of rebleeding that endoscopic and IV PPI therapy is generally considered unnecessary, although oral PPI administration is indicated to initiate ulcer healing. There is some evidence that high-dose oral PPIs are beneficial for peptic ulcer bleeding, but it remains unclear whether such a strategy is appropriate for all endoscopic stigmata and how well oral PPIs compare with their IV counterparts.

When data are lacking regarding the efficacy of a therapy, or perhaps are unconvincing due to clinical trial methodology limitations, it is often useful to conduct an overview of the available literature in order to increase the power of otherwise underpowered studies. Such an undertaking requires the fulfillment of several evidence-based medicine principles. These principles are as follows: that a comprehensive literature search be performed to identify articles that address a predetermined primary outcome; that identified articles are independently reviewed with methodologic quality assessment of each identified article or study; that an independent data-collection process is performed for individual articles; and that investigators agree about the quality, results, and rationale for combining data from individual articles. A qualitative overview in which individual trials are not similar enough for data compilation is called a systematic review. A quantitative systematic review is called a meta-analysis. A rich resource for such overviews is the Cochrane Collaboration. The Cochrane Collaboration is a database of overviews that encompasses many areas of medicine in which data are lacking or a definitive randomized control trial has not been performed.

Cochrane collaboration meta-analysis. During this year's meeting proceedings, Leontiadis and colleagues^[5] presented their meta-analysis contribution to the Cochrane Collaboration evaluating the role of PPIs (both oral and IV) in the treatment of peptic ulcer bleeding. These study authors identified 20 randomized controlled trials (2768 patients) of PPIs vs placebo and H₂-receptor antagonists (H₂RAs) in the treatment of peptic ulcer bleeding. Outcome variables included rebleeding, surgical intervention, and mortality. Taken as a whole, the administration of PPIs (oral and IV) was clinically superior to placebo and H₂RAs for rebleeding (odds ratio [OR]: 0.44, 95% confidence interval [CI]: 0.32-0.60, P < .0001) and the need for surgery (OR: 0.57, 95% CI: 0.40-0.79, P < .001). The number needed to treat (NNT) with PPIs rather than placebo or H₂RAs to prevent 1 episode of rebleeding was 9 and the NNT to prevent 1 surgery was 25. When only studies that included appropriate endoscopic therapy prior to randomization (implying higher-risk ulcer stigmata) were considered, PPI administration was found to have an OR for rebleeding of 0.38 (CI: 0.23-0.61, P < .001) and for surgery of 0.36 (CI: 0.19-0.66, P = .001) compared with placebo or H₂RAs. Intravenous PPIs after appropriate endoscopic therapy was associated with an OR of 0.40 (CI: 0.24-0.67, P < .001) for rebleeding and 0.50 (CI: 0.30-0.84, P = .008) for surgery compared with placebo or H₂RAs.

This meta-analysis lends additional credence to previous studies demonstrating a clinically important benefit resulting from the use of PPIs for peptic ulcer bleeding. Even with this pooled data, however, the demonstration of a mortality benefit with this approach remains unproven and the question of the superiority of 1 form of PPI administration over another (IV vs oral) also remains unanswered. Additional analysis, such as the outcomes associated with individual endoscopic ulcer stigmata or a direct comparison of oral vs IV PPIs, may be useful in determining the optimal route and dose of PPIs in this clinical setting. Overall, however, this is another important piece of evidence that supports the use of PPI therapy as an adjunct in peptic ulcer bleeding.

Nonvariceal upper gastrointestinal bleeding. Sachdev and Gupta^[6] examined the effects of adding IV ranitidine (100-mg bolus followed by 12.5-mg/hr infusion) to IV pantoprazole (80-mg bolus followed by 8-mg/hr infusion) for patients with nonvariceal upper gastrointestinal bleeding. Most patients who experience peptic ulcer rebleeding do so within the first 24-72 hours after the initial bleeding episode. These investigators hypothesized that a more rapid acid suppression might be achieved through the addition of H₂RAs to PPI therapy, perhaps resulting in improved clinical outcomes.

In this small study, they evaluated the intragastric pH of healthy volunteers (n = 16) during a 24-hour infusion of IV PPI and IV PPI + IV H₂RA therapy. They observed that the intragastric pH in the dual-therapy group was significantly greater during the first 6 hours of the infusion period compared with that in the monotherapy group (IV PPI only). The study authors then examined the effects of a 72-hour infusion of IV PPI vs IV PPI + IV H₂RA therapy on clinical outcomes in patients (n = 38) with severe nonvariceal upper gastrointestinal bleeding (systolic blood pressure < 100 mmHg, fresh blood in nasogastric aspirate). Cessation of bleeding with either regimen of acid suppression was approximately 75%. This study was underpowered to demonstrate differences in other outcomes.

Thus, the addition of an IV H₂RA to an IV PPI to achieve a more rapid gastric acid suppression, although theoretically reasonable, does not appear to alter clinical outcome in patients with bleeding PUD and so cannot currently be recommended.

IV PPIs in the ICU*

The routine use of any therapeutic intervention should ideally be limited to indications for which the therapy has clearly demonstrated evidence of improved outcomes. The use of IV PPIs in additional clinical situations continues to engender a great deal of attention with respect to their appropriateness and associated costs. Several studies presented during this year's meeting addressed these issues.

Stress ulcer prophylaxis. In the 1990s, several landmark studies by Cook and colleagues^[7,8] identified patients in critical-care settings at high risk for stress ulcer development (mechanical ventilation > 48 hours and coagulopathy), and demonstrated that acid suppression with H₂RAs decreased the incidence of gastrointestinal bleeding in the intensive care unit (ICU). Due to their more potent acid-suppressing properties, PPIs have been evaluated for stress ulcer prophylaxis in a number of studies. Results from these trials have not been convincing and all are subject to important methodologic limitations largely due to their small size and use of intermediate endpoints (intragastric pH). Because stress ulceration in the ICU is a relatively rare event (0.5%-2.5%) with a mortality of approximately 50%,^[7] additional large trials of stress ulcer prophylaxis are needed to clarify the issue.

Cimetidine is the only FDA-approved acid-suppressing agent for stress ulcer prophylaxis. Smith and Karlstadt^[9] presented the results of a trial conducted in enterally fed patients in the ICU who were believed to be at high risk for stress ulcer bleeding. They compared the intragastric pH achieved with IV pantoprazole (in 6 different dosing regimens) with the recommended infusion dose of cimetidine for stress ulcer prophylaxis (300-mg bolus followed by 50-mg/hr continuous infusion) both before and after initiation of enteral feeds. Intragastric pH was found to increase in the PPI groups at both 8 and 16 hours after initiation of feeds. The intragastric pH in patients treated with cimetidine, however, was found to be lower (more acidic) at both 8 and 16 hours, indicating that perhaps cimetidine is subject to early tachyphylaxis. There also did not appear to be a significant dose-response gradient between the lowest dose of pantoprazole (40 mg once daily) and the highest dose (80 mg thrice daily). Intragastric pH measurements were similar between all dosing regimens of pantoprazole, suggesting that 40 mg once daily is an appropriate dosing schedule.

This study provides additional indirect evidence that PPIs are more effective than H₂RAs for stress ulcer prophylaxis -- but the small size and lack of clinical outcomes data limit the conclusions that can be drawn from this investigation.

Is IV PPI therapy appropriate in the critical-care setting? Mian and colleagues^[10] presented the results of a retrospective chart review examining the clinical indications for IV PPI use at their institution. During a 6-month period, 74 patients in the medical, surgical, and cardiothoracic ICUs of their institution received treatment with IV pantoprazole. Ninety-seven percent of patients had identified risk factors that placed them at high risk for the development of clinically important stress ulcer bleeding. Seventy-three percent of patients received a dose of 40-mg IV pantoprazole once daily and the indication for most of these patients (87%) was stress ulcer prophylaxis. Only 7% of patients treated with IV pantoprazole were treated for bleeding PUD, and these patients received a dose of 8 mg/hr as a continuous infusion. It is interesting to note that 30% of patients in this study received enteral feeding coincident with IV PPI administration, and it is conceivable then that these patients could have been treated with oral PPIs. Adverse events related to the administration of IV pantoprazole were not reported; however, the entire class of PPIs is considered to be very safe, and pantoprazole in particular has been demonstrated to not have important interactions with other pharmaceutical agents commonly encountered in the critical-care setting.

Physician understanding of indications and their prescribing patterns. The increasing use of this novel therapy highlights another important aspect of IV PPI use: physician knowledge of the clinical situations in which these agents may be clinically useful.

White and colleagues^[11] presented the results of a survey evaluating resident physician use and understanding of IV PPI administration. The majority of residents responding (n = 100) agreed that IV PPIs were indicated for active upper gastrointestinal bleeding (91%), EE (75%), complicated GERD (72%), and stress ulcer prophylaxis (58%). More than 50% of the surveyed respondents had used IV PPI therapy for upper gastrointestinal bleeding and stress ulcer prophylaxis. Alarmingly, 47% of the residents believed that IV PPI therapy was indicated for variceal bleeding and 19% thought that this therapy was indicated for active lower gastrointestinal bleeding. Although extending the indications of IV pantoprazole to encompass treatment of active nonvariceal upper gastrointestinal bleeding and prophylaxis of patients at high risk for stress ulcer bleeding is not unreasonable, and makes intuitive sense based on current evidence, clinical benefit with pantoprazole for these indications remains to be proven. Additionally, it is unlikely that any IV PPI would have a beneficial role in patients with variceal bleeding or active lower gastrointestinal bleeding, and therefore such use should be discouraged.

Consequences of inappropriate IV PPI use. The results of a study presented during these meeting proceedings highlight the economic consequences of inappropriate use of IV PPI therapy. Lee and colleagues^[12] performed an observational study of consecutive nonintensive-care patients prescribed IV pantoprazole at their institutions. Use of IV PPI therapy was considered "appropriate" for nothing-by-mouth (NPO) patients with GERD, PUD, or previous use of oral PPIs, for treatment of upper gastrointestinal bleeding, or for stress ulcer prophylaxis in high-risk patients (prolonged mechanical ventilation or coagulopathy). Over a 14-month period, 64 patients received IV PPI therapy. The use of IV PPIs was found to be "appropriate" according to the above indications in only 22% (14/64) of patients. Stress ulcer prophylaxis was the documented or presumed indication for 70% (45/64) of patients receiving IV PPI therapy and was considered "appropriate" in only 13% (6/45). When the indication was upper gastrointestinal bleeding (n = 10), the application was considered "appropriate" in only 30% of patients. Additionally, the doses of IV PPIs used per patient were roughly 3 times the recommended dose for the "appropriate" indications. This inappropriate dosing resulted in a 292% increase in costs attributable to the medication and an excess expenditure of $7766 for the 64 patients, regardless of the "appropriateness" of the IV PPI.

Commentary. All of the trials previously discussed indicate that additional clinical trials and education regarding the use of acid suppression in the critical-care setting as well as improved communication between pharmacy personnel and prescribing physicians are needed for the use of IV PPIs in these clinical situations.

PPIs in the Management of EE

Within the spectrum of GERD, approximately 20% to 40% of patients will have EE, whereas 50% to 70% will have nonerosive reflux disease (NERD). It is well recognized that patients with EE are more likely to develop complications (bleeding, strictures, Barrett's esophagus) than are patients with NERD. It is also well accepted that, as a class, PPIs are the preferred agents for treating patients with EE for both initial healing and maintenance of healing when compared with other antisecretory drugs or placebo.

Treatment of elderly patients with EE. Elderly patients with GERD have been shown to have an even greater risk of GERD-related complications compared with younger patients. ^[13] To assess the efficacy of PPI therapy in this subset of patients, DeVault and colleagues^[14] reviewed data from 2 previous trials comparing pantoprazole with placebo or nizatidine for healing of EE. The 2 trials included 44 elderly (> 65 years of age) patients and 210 younger (< 65 years) patients. Approximately 50% of elderly patients had Grade 2 EE (moderate on the Hetzel-Dent scale) and 50% had Grade 3-4 EE (severe). Compared with younger patients, elderly patients were more likely to have severe EE. Healing rates, however, were similar between the 2 groups. After 8 weeks of pantoprazole therapy, 96% of elderly patients with Grade 2 EE were healed and 76% with Grades 3-4 EE were healed (86% healing at 8 weeks for all grades).

The healing rates for elderly patients treated with pantoprazole were significantly better than for those on placebo or nizatidine, and they mirrored the performance characteristics of pantoprazole in the younger population, indicating that the majority of patients, regardless of age, will do well with standard therapy.

Maintenance of EE healing. As previously mentioned, PPI therapy has been shown to be more effective than H₂RA therapy for maintenance of EE healing. Some patients will relapse, however, even with PPI therapy. Metz and colleagues^[15] reported the results of 2 large studies evaluating the effectiveness of pantoprazole in various daily doses (10 mg, 20 mg, 40 mg) compared with ranitidine (150 mg twice daily) for preventing recurrence of severe EE. After 12 months, patients treated with ranitidine in each study were more likely to have experienced recurrence of severe EE (13%-18%) compared with patients treated with pantoprazole. The 40-mg daily dose of pantoprazole was the most effective maintenance dose among those studied, with 12-month maintenance rates ranging from 78% to 82%.

The results of this study are consistent with previously observed maintenance rates with other PPIs and serve as additional evidence confirming the superiority of PPIs over H₂RAs for maintenance therapy in EE. They also demonstrate that recurrence of EE during maintenance therapy, when it occurs, is less likely to be severe in patients taking PPIs compared with similar patients taking H₂RAs.

Clinical utility of newest PPI for healing and maintenance of EE. Johnson and colleagues^[16,17] evaluated the clinical utility of esomeprazole, the newest PPI, for healing and maintenance of EE in 2 studies presented during these meeting proceedings. The first study examined the therapeutic gain (absolute risk reduction [ARR]) of esomeprazole compared with omeprazole (20 mg) and lansoprazole (30 mg) observed in 4 randomized controlled trials that were published previously.^[18-21] These 4 trials demonstrated a statistically significant improvement for 8-week healing rates of patients with EE treated with esomeprazole compared with the other 2 PPIs. The absolute difference, however, was small and the clinical significance of these differences was debatable.

In the study presented at this meeting, Johnson and colleagues^[16] performed a subgroup analysis and examined the trial participants who had severe (LA Classification C and D) EE. They found that the ARR attributable to esomeprazole compared with the other PPIs in patients with severe EE was significantly greater than that observed when all grades of EE were included in the analysis. That is, esomeprazole appears to be more effective for severe EE compared with omeprazole and lansoprazole. Compared with omeprazole, the ARR with esomeprazole ranged between 10% and 20%, and compared with lansoprazole, the ARR for healing of severe EE was approximately 12%. The reciprocal of the ARR is the NNT. The NNT with esomeprazole to avoid 1 treatment failure with omeprazole or lansoprazole is 5-10 and 8, respectively. This post-hoc analysis, although not the primary objective of the original trials, was adequately powered to demonstrate these differences and indicates that esomeprazole is more effective than at least standard doses of omeprazole and lansoprazole for initial healing of severe EE. These differences appear to be more clinically relevant than those found in previous trials comparing different PPIs for all grades of EE.

The second study presented by Johnson and colleagues^[17] examined the utility of esomeprazole vs lansoprazole (15 mg) for healing maintenance of EE. The primary endpoint was endoscopic remission after 6 months of PPI use. This investigation, like the previous one, indicates that esomeprazole is more effective than lansoprazole for severe EE. However, this study should be examined critically before accepting its conclusions, because there are a number of potential methodologic limitations -- the analysis is a retrospective subgroup analysis and may not be adequately powered; the number of patients with severe EE was relatively small; and measures of statistical significance were not reported for the observed differences between the 2 PPIs. Additionally, it is not clear whether the analysis was reported as intention-to-treat or as a per-protocol analysis, and the follow-up period was only 6 months whereas most maintenance studies to date have been 12-month studies. Last, and perhaps most important, the dose of lansoprazole (15 mg) is lower than what many would consider to be sufficient therapy. It is certainly possible that such treatment differences would not be observed with a 30-mg lansoprazole comparator. Thus, additional data are needed before a definitive claim for superiority of one PPI over another can be made for maintenance therapy in patients with EE.

PPIs in Nocturnal GERD

Another group that is at higher risk for GERD-related complications are patients with nocturnal GERD symptoms.^[22] Patients with nocturnal symptoms often require dose escalations of antisecretory medications. Fass and colleagues^[23] examined the results of 2 PPI regimens in patients with nocturnal heartburn symptoms and reported their results. Two hundred eighty-two patients with nocturnal GERD symptoms resistant to once-daily lansoprazole (30 mg) were randomized to receive either 40 mg of esomeprazole daily or 30 mg lansoprazole twice daily. The treatment period was for 8 weeks and symptom response was patient-derived through daily diary entries.

At the end of the treatment period, a similar response was observed in both treatment arms. Approximately 46% of patients experienced nocturnal GERD symptoms prior to study initiation, and at the end of therapy, only 23% to 26% of patients in each group were reporting nocturnal symptoms. Thus, daily esomeprazole (40 mg) appears to be clinically similar to twice-daily lansoprazole (30 mg) for this difficult-to-treat group of patients, and the choice of therapy will be dictated by cost, availability, and patient preference. It is likely that among the roughly 25% of nonresponders in each group, there are some whose nocturnal symptoms are not due to acid regurgitation and who may in fact have nonulcer dyspepsia or other functional causes of their symptoms. These patients would be candidates for 24-hour pH testing while on PPI therapy in an attempt to determine the degree of acid reflux and correlate reflux episodes with symptoms that could perhaps lead to alternative therapeutic options. The other important point that this study raises is that escalating to a twice-daily regimen in patients with resistant GERD symptoms is the preferred mechanism of dose increase, rather than increasing a once-daily dose.

PPIs in the Management of NSAID-Associated Dyspepsia*

In addition to GERD and EE, PPIs are commonly used to treat dyspepsia symptoms. Dyspepsia (broadly defined as epigastric abdominal pain) is quite common in patients taking nonsteroidal antiinflammatory drugs (NSAIDs) and is a leading cause of noncompliance or discontinuation of these medications. NSAIDs are also a leading cause of PUD, and dyspeptic symptoms in NSAID-taking patients may be associated with active ulcer disease or simply gastrointestinal intolerance of the NSAID. Patients with a previous history of NSAID-induced ulcers are at an increased risk for future ulcer formation and are also more likely to experience NSAID-associated dyspepsia. Unfortunately, for many patients, NSAIDs are an important and beneficial medication for chronic pain and they are reluctant to discontinue their use, even after an ulcer complication. Misoprostol, PPIs, and cyclooxygenase (COX)-2 selective NSAIDs have been shown to decrease the incidence of dyspeptic symptoms and NSAID-associated ulcer complications in several randomized controlled trials, but have not been compared directly for their effects on the development of NSAID-associated dyspepsia.^[24-26]

Chey and colleagues^[27] evaluated the effect of lansoprazole (15 and 30 mg daily), misoprostol (200 micrograms 4 times daily), and placebo on NSAID-associated dyspepsia in a large group of patients with a history of NSAID-induced ulcers who continued taking NSAIDs. The endpoints consisted of a composite of gastrointestinal symptoms (abdominal pain, heartburn, belching, fullness/bloating/early satiety, and abdominal distension) as well as the individual symptoms of abdominal pain and heartburn. At the completion of the 12-week trial, lansoprazole in both doses was associated with freedom from gastrointestinal composite symptoms in about 66% of patients, abdominal pain in about 77%, and heartburn in about 87%. Symptom scores were similar between patients taking placebo and misoprostol, although fewer patients taking misoprostol reported abdominal pain. Lansoprazole was significantly more effective than placebo for all measured endpoints and was more effective than misoprostol for composite gastrointestinal and heartburn symptoms. It is important to recognize that misoprostol is historically difficult for patients to tolerate due to the frequent development of diarrhea and abdominal cramping, although lower doses than that used in this study have been shown to be better tolerated and still be gastrointestinal-protective. Caution must also be exercised when using misoprostol in women of childbearing age.

The take-home message from this study is that in patients with a history of NSAID-associated gastrointestinal symptoms who continue taking NSAIDs, the addition of a once-daily dose of lansoprazole reduces the risk for future dyspeptic symptoms and is more effective than misoprostol and placebo. This represents an alternative to the expensive COX-2 selective NSAIDs and should be considered in appropriate patients. These observed benefits are likely a class effect, but additional studies (similar to this current one) evaluating the other available NSAIDs are needed.

Concluding Remarks

We continue to gain important insights regarding the most appropriate uses of PPIs for a variety of acid-peptic disorders. Based on presentations during this year's ACG meeting, the clinical evidence to support the use of IV PPIs in patients with bleeding PUD (especially after endoscopic therapy) and stress ulcer prophylaxis continues to increase. It is important, however, to use these medications judiciously, understanding their effects and associated costs. A close integration with the pharmacy is advisable and additional large-scale outcomes research trials are needed. Individual comparisons between different PPIs for the treatment of EE are slowly accumulating and indicate that there may be some advantages of the newer PPIs for more severe disease. The clinical relevance of these differences is likely to be moderated by cost and availability.

As our experience with PPIs continues to increase, it is likely that many of the current questions surrounding the clinical applications and cost-effectiveness of these agents for acid-mediated conditions will be clarified and that current alternative uses of PPIs will become well-established, evidence-based medicine practices in the future.

*This section mentions off-label uses for some medications. These may describe clinical uses for medications that have not been approved by the FDA.

Disclaimer: The opinions and assertions contained herein are the sole views of the author and should not be construed as official or as representing the views of the US Navy or Department of Defense.

References

1. Laine L, Peterson WL. Bleeding peptic ulcer. N Engl J Med. 1994;331:717-727.
2. Cook DJ. Endoscopic therapy for acute nonvariceal upper gastrointestinal hemorrhage: a meta-analysis. Gastroenterology. 1992;102:139-148.
3. Lin HJ, Lo WC, Lee FY, et al. A prospective randomized comparative trial showing that omeprazole prevents rebleeding in patients with bleeding peptic ulcer after successful endoscopic therapy. Arch Intern Med. 1998;158:54-58.
4. Lau JYW, Sung JJY, Lee KKC, et al. Effect of intravenous omeprazole on recurrent bleeding after endoscopic treatment of bleeding peptic ulcers. N Engl J Med. 2000;343:310-316.
5. Leontiadis GI, McIntyre L, Sharma VK, et al. Proton pump inhibitors (PPI) in the treatment of peptic ulcer bleeding: a Cochrane Collaboration meta-analysis of randomized controlled trials. Am J Gastroenterol. 2003;98:S49. [Abstract #142]
6. Sachdev GK, Gupta SK. Effect of addition of IV ranitidine to IV pantoprazole on 24 hour intragastric pH and its role in control of non-variceal upper GI bleed. Am J Gastroenterol. 2003;98:S50. [Abstract #145]
7. Cook DJ, Fuller HD, Guyatt GH, et al. Risk factors for gastrointestinal bleeding in critically ill patients. Canadian critical care trials group. N Engl J Med. 1994;330:377-381.
8. Cook DJ, Reeve BK, et al. Stress ulcer prophylaxis in critically ill patients. Resolving discordant meta-analyses. JAMA. 1996;275:308-314.
9. Smith JS, Karlstadt RG. Rapid gastric pH improvement from NPO to enteral-fed period with intermittent intravenous (IV) pantoprazole (P) vs. continually infused cimetidine (C). Am J Gastroenterol. 2003;98:S41. [Abstract #119]
10. Mian NA, White KT, Michele L, et al. Is the use of IV pantoprazole appropriate in the critical care setting? Am J Gastroenterol. 2003;98:S234. [Abstract #707]
11. White KT, Storch IM, Mullin GE. Intravenous pantoprazole: physician understanding of indications and their prescribing patterns leading to hospital restrictions. Am J Gastroenterol. 2003;98:S231. [Abstract #697]
12. Lee EY, Brotz C, Greenwald D. Cost and characteristics of intravenous proton pump inhibitor use and misuse in a tertiary care hospital. Am J Gastroenterol. 2003;98:S54. [Abstract #154]
13. Collen MJ, Abdulian JD, Chen YK. Gastroesophageal reflux disease in the elderly: more severe disease that requires aggressive therapy. Am J Gastroenterol. 1995;90:1053-1057.
14. DeVault KR, Lynn RB, Bochenek WJ, et al. Successful treatment of elderly patients with erosive esophagitis (EE) using pantoprazole 40 mg. Am J Gastroenterol. 2003;98:S3. [Abstract #7]
15. Metz DC, Fraga P, Mack ME, et al. Patients with erosive esophagitis relapse less frequently and to lower grades after treatment with pantoprazole v ranitidine. Am J Gastroenterol. 2003;98:S29. [Abstract #84]
16. Johnson DA, Hoyle PE, Traxler BM, et al. An evidence- based approach provides a quantitative assessment of the efficacy of esomeprazole for healing of erosive esophagitis based on disease severity. Am J Gastroenterol. 2003;98:S16. [Abstract # 44]
17. Johnson DA, Hoyle PE, Traxler BM, et al. Esomeprazole as maintenance therapy in erosive esophagitis: a quantitative assessment of efficacy using an evidence-based approach. Am J Gastroenterol. 2003;98:S34. [Abstract # 97]
18. Richter JE, Kahrilas PJ, Johanson J, et al. Efficacy and safety of esomeprazole compared with omeprazole in GERD patients with erosive esophagitis: a randomized controlled trial. Am J Gastroenterol. 2001;96:656-665.
19. Kahrilas PJ, Falk GW, Johnson DA, et al. Esomeprazole improves healing and symptom resolution as compared with omeprazole in reflux oesophagitis patients: a randomized controlled trial. The Esomeprazole Study Investigators. Aliment Pharmacol Ther. 2000;14:1249-1258.
20. Nexium [package insert]. Wilmington, DE: AstraZeneca LP; 2003.
21. Castell DO, Kahrilas PJ, Richter JE, et al. Esomeprazole (40 mg) compared with lansoprazole (30 mg) in the treatment of erosive esophagitis. Am J Gastroenterol. 2002;97:575-583.
22. Lagergren J, Bergstom R, Lindgren A, et al. Symptomatic gastroesophageal reflux as a risk factor for esophageal adenocarcinoma. N Engl J Med. 1999;18:340:825-831.
23. Fass R, Traxler BM, Thomas S, et al. Once daily esomeprazole and twice daily lansoprazole for nocturnal heartburn in patients with GERD. Am J Gastroenterol. 2003;98:S36. [Abstract #102]
24. Silverstein FE, Graham DY, Senior JR, et al. Misoprostol reduces serious gastrointestinal complications in patients with rheumatoid arthritis receiving nonsteroidal anti-inflammatory drugs. A randomized, double-blind, placebo-controlled trial. Ann Intern Med. 1995;123:241-249.
25. Yeomans ND, Tulassay Z, Juhasz L, et al. A comparison of omeprazole with ranitidine for ulcers associated with nonsteroidal antiinflammatory drugs. Acid Suppression Trial: Ranitidine versus Omeprazole for NSAID-associated Ulcer Treatment (ASTRONAUT) Study Group. N Engl J Med. 1998;338:719-726.
26. Bensen WG, Zhao SZ, Burke TA, et al. Upper gastrointestinal tolerability of celecoxib, a COX-2 specific inhibitor, compared to naproxen and placebo. J Rheumatol. 2000;27:1876-1883.
27. Chey WD, Lukasik N, Huang B, et al. Lansoprazole effectively reduces the risk of GI symptom occurrence in ulcer free patients who continue chronic NSAID use. Am J Gastroenterol. 2003;98:S296. [Abstract #888]