Long-term NSAID Use and Acid-Related GI Complications: Improving the Outcomes

Nonsteroidal anti-inflammatory drug (NSAID) gastropathy and upper gastrointestinal (GI) bleeding from peptic ulcers remain frequent and clinically important events to both gastroenterologists and primary care physicians. While NSAID use remains ubiquitous, the risk of clinically important side effects is 2% to 5% for complicated or symptomatic ulcer disease and 40% or more for NSAID-related dyspepsia. The latter is often amenable to antisecretory therapy with histamine 2-receptor antagonists (H2RAs) or proton-pump inhibitors (PPIs), but the optimal management approach is unclear. It appears that PPIs are superior to H2RAs in this regard, but whether one PPI is superior to others remains to be determined.

Complicated NSAID gastropathy is less frequent than NSAID-induced dyspepsia, but it is a potentially life-threatening complication of NSAID use and thus a very important clinical consideration. The risk of serious GI toxicity related to NSAIDs appears to increase as one moves from low-dose aspirin to cyclooxygenase 2 (COX-2) inhibitors (coxibs) to coxibs plus aspirin or nonselective NSAIDs to these older NSAIDs plus aspirin. This last group of patients is numerous, as many older patients in the United States use both over-the-counter or prescription traditional nonselective NSAIDs plus cardioprotective doses of aspirin. Such patients are at the highest risk for ulcer disease and its complications, but this fact is underappreciated by clinicians.

Proton-pump inhibitor co-therapy with either nonselective NSAIDs or coxibs appears to lower the risk of ulceration from these agents and is an effective prophylactic strategy. This strategy is unlikely to prevent the recently described small-bowel lesions associated with nonselective NSAIDs, although the clinical relevance of such lesions remains controversial. Coxibs appear not to increase the risk of these small-bowel findings. Whether the overall safest strategy in preventing NSAID complications is a nonselective NSAID plus a PPI or a coxib alone has not yet been studied, and the choice of approaches remains at the discretion of the treating physician. In a very high-risk patient requiring NSAIDs, the combination of a PPI plus a coxib is likely the safest strategy.

Bleeding peptic ulcers secondary to NSAIDs and/or infection with Helicobacter pylori continue to be a prevalent and clinically important event. Ulcer healing occurs in most with either discontinuation of the NSAID and/or eradication of the infection. However, prevention of rebleeding while awaiting ulcer healing is not achieved by such a strategy. Endoscopic therapy of the bleeding ulcer is effective in decreasing rebleeding from 50% or more to 20% or less. The addition of a PPI to endoscopic therapy decreases rebleeding even further, down to 5% or less. The optimal formulation of PPI and dose is as yet unknown. The best evidence supports the use of an intravenous (IV) bolus followed by a continuous infusion, although there is also evidence that higher doses of oral agents are effective as well. Currently, most experts recommend a continuous infusion, but this may not be a universal approach.

The issues related to NSAID use and peptic ulcer bleeding continue to be of clinical relevance, and much remains to be learned regarding these disease states. However, the data presented at Digestive Disease Week (DDW) 2004 and described herein have contributed substantial information in these arenas and clarified some of the ongoing controversies.

In patients who are chronic users of NSAIDs, preventing the development of symptomatic ulcers and ulcer complications is quickly becoming a priority for clinicians. This important issue was the focus of discussions by Jay Goldstein, MD, University of Illinois, Chicago, and Ian Gralnek, MD, MSHS, David Geffen School of Medicine at the University of California at Los Angeles. The experiences and advice of Drs. Goldstein and Gralnek were part of an industry-supported continuing medical education (CME) activity entitled, "Acid-Related GI Complications: Can We Improve Outcomes?" and are summarized below.

According to Dr. Goldstein, 1% to 2% of people on NSAIDs will present with an upper GI bleed, perforation, or symptomatic ulcer. This risk increases nearly eightfold if the patient is taking 2 nonsteroidal agents, even if one of those agents is low-dose aspirin. This risk, Dr. Goldstein said, is constant over time with no adaptation to NSAIDs. Older patients who have a history of ulcers are at higher risk of GI complications and tend to be on higher doses or multiple NSAIDs, take corticosteroids, and have major illnesses. Risk of bleeding in these older patients increases exponentially with the addition of each risk factor.

In the Misoprostol Ulcer Complications Outcomes Safety Assessment (MUCOSA) trial, which involved 8800 patients, misoprostol taken with NSAIDs reduced upper GI complications by 50% compared with NSAIDs alone. Misoprostol has been shown to help reduce the rate of gastric ulcers (GU) and duodenal ulcers (DU) as well as the complications of ulcers; however, there are a number of side effects associated with its use, and it must be taken 3 times a day. Histamine 2 blockers can help protect against DU, but not gastric ulcers, and the benefit of H2 blockers in preventing ulcer complications is not known. Proton-pump inhibitors, on the other hand, have been proven to protect against both GU and DU, and are likely to help prevent ulcer complications. Another study cited by Dr. Goldstein[1] compared ranitidine (an H2 blocker) with omeprazole (a PPI) in patients who had a healed ulcer but continued to take NSAIDs for 6 additional months. This population was at high risk of ulcer recurrence, and those taking ranitidine had a recurrence rate of 16.3% and 4.2% for GU and DU, respectively, vs 5.2% and 0.5%, respectively, for those taking omeprazole.

In a similar study by Scheiman and colleagues, patients with healed ulcers were randomized to placebo (NSAIDs alone, including COX-2 types) or NSAIDs plus either 20 mg or 40 mg of esomeprazole daily. At the end of 6 months, ulcers had recurred in 12.3% of the NSAIDs-alone group vs 4.4% and 5.2% in the 20-mg and 40-mg esomeprazole-treated patients, respectively.

In the Vioxx Gastrointestinal Outcomes Research Study (VIGOR) trials, according to Dr. Goldstein, results showed that COX-2 inhibitors reduced the rate of ulcers and ulcer complications by 50%, demonstrating the efficacy of this class of drug in a non-aspirin-using population.

Aspirin is toxic in the upper GI tract, and despite common belief, it does not matter how aspirin is administered as its GI toxicity is based on a systemic effect, not its topical effect. Thus, all forms of aspirin can cause bleeding based on aspirin's systemic effect. Aspirin used alone or with celecoxib can cause similar bleeding rates, but aspirin is even more toxic when used together with a traditional NSAID. The combination of naproxen and aspirin produced a 27% ulcer rate in subjects. This rate was lowered when aspirin was combined instead with celecoxib, but it was not lowered to the same level as placebo. Thus, Dr. Goldstein suggested that there is some adverse effect when switching from aspirin to celecoxib plus aspirin.

In high-risk patients (those over 65 with a prior GI bleed), the rate of bleeding when taking NSAIDs is almost 29%, but can be reduced to 16% with a COX-2 inhibitor. However, that is still a high percentage. In a study of patients with healed ulcers[2], patients were treated with either celecoxib or the combination of omeprazole and diclofenac. Results showed that the rate of rebleeding was reduced to 4.9% in the celecoxib group and to 6.4% in the omeprazole-plus-diclofenac group.

In the lower GI tract, bleeding (at a rate of 1%) can be reduced by 50% through the use of COX-2-type drugs, as cyclooxygenase 1 (COX-1) is expressed in the small bowel. The Celecoxib Long-term Arthritis Safety Study (CLASS) trial demonstrated that slow bleeds in the small bowel are less frequent with coxibs. In a trial with healthy volunteers[3], subjects were randomized to placebo, naproxen plus omeprazole, or celecoxib. The naproxen-plus-omeprazole group had an average of 3 lesions or mucosal breaks at the end of the study compared with 0.32 in the celecoxib group; 55% of the subjects taking the traditional nonsteroidal had mucosal breaks. NSAIDs affect not only the upper GI tract but also the small bowel and the colon.

Dr. Goldstein noted that physicians should be aware of the GI toxicity of aspirin and nonselective NSAIDs, as well as the increased risk in certain populations. These risks can be addressed by the use of protective therapies and COX-2-specific inhibitors.

Dr. Gralnek discussed some of the newer strategies being used to prevent upper GI bleeding, rebleeding, and, in particular, the role of medical therapy. He suggested that patients who have an acute upper GI bleed and peptic ulcer disease (PUD) should be put on PPIs, have their aspirin and NSAIDs (including coxibs) suspended, and be investigated for H pylori.

Dr. Gralnek said that PPIs, rather than H2 blockers, should be used to treat bleeding, as it is known that pH can rise to above 6 with IV PPI administration; pH levels above 6 allow for better platelet aggregation and inactivate pepsin, which can facilitate clot formation and retention, theoretically reducing the risk of rebleeding. In a meta-analysis of the use of IV H2 blockers vs placebo in bleeding, results showed that there was no difference in rebleeding, the need for surgery, or mortality rates between the 2 treatment arms.

Dr. Gralnek also said that there is a role for high-dose oral PPIs in the context of patients who present with upper GI bleeding. In a study of patients with upper GI bleeds from ulcers administered 40 mg omeprazole twice a day vs placebo[4], there was a significant improvement in PPI-treated patients. Rates of rebleeding were 10.9% vs 36.4%, rates of surgery were 7.3% vs 23.6%, and rates of transfusion were 29.1% vs 70.9% for the high-dose PPI arm and the placebo arm, respectively (P < .001 for all). There was no significant difference in the rate of mortality.

Dr. Gralnek added that in a follow-up to this study, patients who had a GU or DU with high-risk stigmata received endoscopic hemostasis and were randomized to high-dose PPI or placebo. The rates of rebleeding and the need for transfusion were significantly lower in the high-dose PPI-treated patients vs placebo (7% vs 21%, respectively, for rebleeding; 35% vs 73%, respectively, for transfusions), but there were no differences in the need for surgery or in mortality.

Intravenous PPI was compared with IV cimetidine in a similar group of patients. After endoscopic hemostasis, patients were randomized to receive a bolus and continuous infusion of either omeprazole or cimetidine for 72 hours. At day 3, there was no rebleeding in the PPI group compared with 16% in the cimetidine group. At day 14, rebleeding had occurred in 4% of the PPI-treated patients and 24% of the H2-blocker treated patients, while mortality was 0% and 4%, respectively, according to Dr. Gralnek. In a follow-up study comparing a PPI with placebo[5], the rate of bleeding at day 3 was 4.2% vs 20%, respectively, and 5.8% vs 21.7%, respectively, at the end of 7 days. A further study indicated that IV PPI with endoscopic hemostasis is more effective than IV PPI alone. Rebleeding after 30 days, reported Dr. Gralnek, was 1% in the combination group vs 9% for those treated with IV PPI alone.

Dr. Gralnek concluded that patients should start on PPIs prior to endoscopy, but it is up to the treating physician to decide if they should get high-dose oral or IV PPIs. Patients should be treated with endoscopic hemostasis and should be given IV PPIs for 72 hours postendoscopy if they have high-risk stigmata.

Home | Articles | Writing Credits | Writing Reviews
Teaching | Reviews on Teaching | Contact