P&T News: March/April 2003

Cyclo-Oxygenase-2 Selective Inhibitors Versus Proton-Pump Inhibitors: Efficacy in Reducing the Incidence of Adverse Gastrointestinal Events

Ryan Jacobsen, Pharm.D., Primary Care Specialized Resident, Department of Pharmaceutical Care
Peer Review Status: Internally Peer Reviewed by Joel V. Weinstock, Professor and Head, Division of Gastroenterology -- Hepatology, Department of Internal Medicine


Nonsteroidal anti-inflammatory drugs (NSAIDs) such as ibuprofen and naproxen are some of the most commonly used medications in the United States for both acute and chronic pain. While NSAIDs are extremely effective analgesic and anti-inflammatory agents, they are plagued by a high incidence of adverse gastrointestinal (GI) effects. In fact, NSAIDs are reported to increase the risk of serious upper GI events approximately four-fold.1 NSAID use among elderly patients in the United States is responsible for over 41,000 hospitalizations and 3,300 deaths annually.2 It then comes as no surprise that a treatment approach that minimizes or prevents NSAID-related adverse GI events would be a desirable therapeutic goal. This article reviews two therapeutic options for decreasing adverse GI events in high-risk patients who require long-term therapy with NSAIDs.

Background

NSAIDs inhibit a family of enzymes known as the cyclo-oxygenases, which are directly responsible for both the efficacy and toxicity of nonsteroidal agents. Cyclo-oxygenases play a predominant role in the formation of pro-inflammatory prostaglandins at sites of injury and of muco-protective prostaglandins in the GI tract. The family of cyclo-oxygenases can be divided into subtypes, cyclo-oxygenase-1 (COX-1) and cyclo-oxygenase-2 (COX-2). The anti-inflammatory activity of the NSAIDs is mediated through inhibition of COX-2, while the toxic effects are a result of COX-1 inhibition in the gastrointestinal mucosa. COX-1 is constitutively expressed in the GI tract, while COX-2 expression is increased at sites of inflammation. The identification of these two isoforms led to the theory that selective inhibition of COX-2 would effectively attenuate the production of pro-inflammatory prostaglandins while sparing the beneficial prostaglandins produced by COX-1 in the GI tract. Therefore, several agents that selectively inhibit COX-2 have been developed, including celecoxib (Celebrex®), rofecoxib (Vioxx®), and valdecoxib (Bextra®).

An alternative to choosing agents that selectively inhibit COX-2 is gastric acid suppression therapy with proton pump inhibitors (PPIs). Agents such as lansoprazole (Prevacid®) and omeprazole (Prilosec®) are effective in preventing NSAID-related peptic ulcers. The physiologic suppression of acid secretion is thought to attenuate the deleterious effects of acid on the GI wall in the face of decreased muco-protective prostaglandin production resulting from inhibition of COX-1 by traditional non-selective NSAIDs.

Clinicians are often faced with the difficult decision of selecting appropriate analgesic or anti-inflammatory therapy when caring for patients at high risk for adverse gastrointestinal events such as bleeding or ulceration. Factors that place patients at high risk include advanced age (over 65 years old), a history of GI bleeding or peptic ulcer disease, multiple or high-dose NSAIDs, and anticoagulant or corticosteroid use. Given the advent of the COX-2 selective inhibitors celecoxib, rofecoxib, and valdecoxib, the treatment of chronic pain and inflammation has experienced a shift over the past five years from non-selective-based NSAID therapy to COX-2 selective therapy. Reasons for this change are multifactorial but rest predominantly on data supporting a decreased incidence of adverse GI events such as gastric and duodenal ulcerations in patients receiving COX-2 selective agents versus non-selective NSAIDs.3,4 While celecoxib and rofecoxib are effective at preventing GI adverse events, their widespread use (the 20th and 24th most prescribed medications in the US in 2001, respectively) has prompted considerable attention among different sectors of health care, including cardiologists and managed care providers.5 Cardiologists have expressed concern about increased cardiovascular events due to a lack of significant inhibitions of platelet aggregation because of selective COX-2 inhibition, while managed care organizations are concerned about increased medication costs. To add to the expense of COX-2 selective inhibitors, patients who receive COX-2 selective therapy because of an increased risk of adverse GI events with NSAIDs are often also given PPIs for various comorbidities (e.g., gastroesophageal reflux disease). Lansoprazole and omeprazole, the 14th and 18th most prescribed medications in 2001, respectively, comprise the majority of prescriptions in this class.5

COX-2 Selective Inhibitors vs.PPIS: Which is More Effective?

It is indeed difficult to compare two completely different classes of medications when assessing GI risk. A number of factors (e.g., GI risk factors, comorbidities, effectiveness of previous treatment, and cost-effectiveness) come into play and must be considered individually for each patient.

Cyclooxygenase-2 Selective Inhibitors

Use with Aspirin
Large randomized, controlled trials support the use of celecoxib and rofecoxib to decrease the incidence of adverse GI events in high-risk patients. The Celecoxib Long-term Arthritis Safety Study (CLASS) compared celecoxib with the non-selective NSAIDs ibuprofen and diclofenac with regard to the incidence of adverse GI events.3 The study permitted aspirin <325 mg/day for cardioprotection. In patients receiving low-dose aspirin, there was no significant difference in adverse GI events (ulcer complications and symptomatic ulcers, P=0.49) between treatment groups. However, in non-aspirin users there was a significant 65% reduction in complicated events with celecoxib (P=0.04), and a 52% reduction in complicated events and symptomatic ulcers (P=0.02) when compared to ibuprofen and diclofenac. Thus, while celecoxib may be effective for decreasing GI complications, its usefulness for minimizing GI risk in aspirin users is questionable. Non-selective NSAID therapy may be appropriate for chronic pain management in aspirin users given that suitable GI prophylactic measures are utilized in high-risk patients.

Use without Aspirin
The Vioxx Gastrointestinal Outcomes Research (VIGOR) trial compared rofecoxib to the non-selective NSAID naproxen and subsequently concluded that rofecoxib was associated with fewer adverse GI events.4 Specifically, the study revealed a relative risk reduction for clinical upper GI events of 50% and a 60% reduction in complicated events with rofecoxib (P<0.001 and P=0.005, respectively). This trial did not include patients on aspirin therapy. Therefore, if a patient does not require cardiovascular protection with aspirin, COX-2 selective inhibitors may be a reasonable choice to help avoid GI adverse events.

Proton-Pump Inhibitors

Compared to Histamine-2 Receptor Antagonists
In long-term NSAID users, both lansoprazole and omeprazole have been shown to be effective for the healing and prevention of NSAID-induced ulcers.6,7,8 Lansoprazole is FDA approved for this indication.9 PPIs are superior in efficacy to both placebo and histamine-2 receptor antagonists (H2RAs) such as ranitidine in preventing NSAID-related GI events. Yeomans et al performed an international double-blind trial with 541 patients requiring chronic NSAID therapy who had a history of NSAID-induced GI lesions.6 Patients were randomized to omeprazole 20 mg daily or 40 mg daily, or ranitidine 150 mg twice daily. Primary endpoints for the study were healing of GI lesions and maintenance of remission. Both doses of omeprazole (20 mg and 40 mg daily) were statistically superior to ranitidine with healing rates at eight weeks of 80%, 79%, and 63%, respectively (P<0.001 for omeprazole 20 mg vs. ranitidine and P=0.001 for omeprazole 40 mg vs. ranitidine).

Compared to Misoprostol
PPIs are considered to have comparable efficacy to misoprostol but with fewer associated adverse effects. Misoprostol, a prostaglandin analogue, was the first FDA- approved medication in the US for reducing the risk of NSAID-related GI complications.10 However, over one-third of patients prescribed misoprostol at effective doses (200 mcg four times daily) experience diarrhea and discontinue therapy. Graham et al conducted a prospective, double-blind, placebo-controlled trial among 537 long-term NSAID users who had a history of endoscopically confirmed gastric ulcer.8 The study was designed to compare misoprostol 200 mcg four times daily, lansoprazole 15 mg daily or 30 mg daily, and placebo in regards to preventing recurrence of gastric ulcer. At 12 weeks following randomization, 51% (95% CI, 41.1-61.3%), 93% (95% CI, 87.2-97.9%), 80% (95% CI, 72.5-87.3%), and 82% (95% CI, 75-89.6%) of patients receiving placebo, misoprostol, lansoprazole 15 mg and lansoprazole 30 mg remained ulcer free, respectively. The authors concluded that when compliance and adverse effects are taken into account, both doses of lansoprazole and misoprostol are equally effective. Given the ease of once daily dosing and fewer associated adverse effects, lansoprazole was identified as the preferred agent. Furthermore, lansoprazole reduces the risk for recurrence of ulcer complications (perforation, bleeding, or obstruction) in patients receiving long-term, low-dose aspirin (P=0.008 compared to placebo).11 Given these data, PPIs are the agents of choice over prostaglandin analogues and H2RAs for treating and preventing NSAID-induced ulcers and complications in patients requiring long-term NSAID therapy.

COX-2 Selective Inhibitors vs. PPIS: Clinical Trials

There are limited data regarding the incidence of adverse GI events when comparing the use of PPIs plus NSAIDs versus COX-2 selective inhibitors in head-to-head trials. However, two small trials, one published in full manuscript form and one published in abstract form, provide insight into the comparative efficacy of COX-2 selective inhibitors versus PPIs plus traditional NSAIDs. Chan et al compared celecoxib (200 mg twice daily) with diclofenac plus omeprazole (75 mg twice daily and 20 mg daily, respectively) in 287 NSAID users for the secondary prevention of ulcer bleeding.12 Subjects were followed for six months for recurrence of endoscopically-confirmed ulcer bleeding. The celecoxib-treated patients had a 4.9% (95% confidence interval, 3.1 to 6.7) six-month probability of recurrence compared to 6.4% (95% confidence interval, 4.3 to 8.4) in the omeprazole group. The authors concluded that treatment with celecoxib was as effective as diclofenac plus omeprazole for the secondary prevention of ulcer bleeding.

Lai et al conducted a small, randomized trial comparing celecoxib (200 to 400 mg daily) with naproxen, a non-selective NSAID, plus lansoprazole (500 to 750 mg daily and 30 mg daily, respectively) in 115 patients with a history of NSAID-related peptic ulcer bleeding.13 No significant difference in recurrence of ulcer complications (3.4% vs. 3.5%, respectively; P>0.05) was found between the two treatment regimens. While these studies provide some evidence for comparable efficacy between traditional NSAIDs plus PPIs versus COX-2 selective inhibitors for reducing the incidence of adverse GI events compared to NSAIDs alone, it is important to keep in mind that these studies were not placebo-controlled and they enrolled only a small number of patients. Therefore, it may be premature to extrapolate these results to the population at large.

Cost-effective Therapy

A number of important factors must be considered when attempting to select the most cost-effective therapy. Some of these factors were reviewed by El-Serag et al and include the comparative cost and effectiveness of the various regimens, the risk factors present in the intended population, and the costs associated with GI complications. The costs vary greatly depending on the institution and patient population served (i.e., high-risk vs. low-risk).14 Currently, given preliminary data suggesting comparable efficacy between the COX-2 selective inhibitors and PPIs plus non-selective NSAIDs, one is faced with looking closely at cost. Based on UIHC acquisition pricing, it appears that PPI-based therapy with a non-selective NSAID is the more cost-effective alternative (see Table 1). Additionally, one must take into account individual patient factors such as concomitant aspirin use, which may make PPIs an even more attractive option, when compared to the use of COX-2 selective inhibitors.

Summary

A treatment regimen consisting of a COX-2 selective inhibitor or a regimen consisting of a PPI plus a non-selective NSAID safely and effectively reduces the risk of adverse GI events compared to the use of traditional NSAIDs alone. While it is not known if adding a COX-2 selective inhibitor to the regimen of those patients receiving a PPI further reduces risk for adverse GI outcomes, using both a COX-2 selective agent and a PPI is very expensive for both the patient and the hospital. Therefore, in patients already receiving a PPI for another indication (e.g., GERD), it is appropriate to prescribe the PPI plus non-selective NSAID combination rather than adding a COX-2 selective inhibitor.

Prescribers are encouraged to consider the following when selecting the most safe and effective therapy for their patients:

Table 1. Cost Comparison of Various Regimens Designed to Reduce the Risk of Gastrointestinal Adverse Effects.

Regimen

Monthly Cost*

COX-2 Selective NSAIDs

Rofecoxib (Vioxx®) 12.5 mg QD

$86

Rofecoxib 25 mg QD

$86

Celecoxib (Celebrex®) 100 mg BID

$106

Celecoxib 200 mg QD

$86

Celecoxib 200 mg BID

$173

Proton Pump Inhibitor (PPI)

Lansoprazole (Prevacid®) 30 mg QD

$139

Nonselective NSAID plus PPI

Naproxen (Naprosyn®) 500 mg BID ($59) plus
Lansoprazole 30 mg QD ($139)

$198

Salsalate (Disalcid®) 750 mg BID ($16) plus
Lansoprazole 30 mg QD ($139)

$155

Ibuprofen (Motrin®) 800 mg TID ($28) plus
Lansoprazole 30 mg QD ($139)

$167

Sulindac (Clinoril®) 200 mg BID ($73) plus
Lansoprazole 30 mg QD ($139)

$212

COX-2 Selective NSAID plus PPI

Rofecoxib 25 mg QD plus
Lansoprazole 30 mg QD

$225

Celecoxib 200 mg QD plus
Lansoprazole 30 mg QD

$225

Celecoxib 100 mg BID plus
Lansoprazole 30 mg QD

$245

*These are current average wholesale price (AWP) acquisition costs; price to the patient may vary depending upon individual pharmacy acquisition costs and pricing policies.

References

1. Am J Med 2001;110(1A):33S-7S.
2. Health Aff (Millwood) 1990;9:114-22.
3. JAMA 2000;284:1247-55.
4. N Engl J Med 2000;343:1520-8.
5. http://www.rxlist.com/top200.htm.
6. N Engl J Med 1998;338:719-26.
7. N Engl J Med 1998;338:727-34.
8. Arch Intern Med 2002;162:169-75.
9. Prevacid‚ (lansoprazole) Full Prescribing Information. 2002 TAP Pharmaceutical Products Inc.
10. Ann Intern Med 1995;123:241-9.
11. N Engl J Med 2002;346:2033-8.
12. N Engl J Med 2002;347:2104-10.
13. Gastroenterology 2001;120(suppl 1):A-104(abstract).
14. Arch Intern Med 2002;162:2105-10.

 

Title Page