BOSTON ( Reuters ) -- ASA and related drugs kill almost as many people every year as AIDS and are responsible for a "silent epidemic," researchers say in an article published today.
Ulcers caused by such drugs kill about 16,500 people in the United States each year, they report in the New England Journal of Medicine. AIDS killed 16,685 Americans in 1997, according to the U.S. National Centre for Health Statistics.
Dr. M. Michael Wolfe of the Boston University School of Medicine said doctors have long known that nonsteroidal anti-inflammatory drugs, or NSAIDs, cause ulcers and other stomach problems in a small fraction of patients.
But because they are used so widely -- about 26 billion tablets are consumed in the U.S. annually -- the number of deaths caused by the drugs is large, Wolfe said in a statement.
"We estimate that approximately 16,500 deaths occur annually in the U.S. from ulcer-related complications associated with their use," he said.
If those deaths were given their own category, the report states, the effects of NSAJD drugs "would constitute the 15th most common cause of death in the United States."
NSAIDs include acetylsalicylic acid ( ASA, sold generically and as Bayer's Aspirin ), ibuprofen, and the active ingredients in Aleve, Naprox, Voltaren and Indocin.
"This is, in many ways, a silent epidemic because gastrointestinal complications are not preceded by any warning signs in many individuals," Wolfe said.
NSAID-Induced GI Complications: What's New?
Identification and prevention of adverse drug reactions (ADRs) is a major focus in health care today. Based on the results of a recent meta-analysis (Lazarou, 1998), it was estimated that 2,216,000 hospitalized patients in the United States experienced a serious adverse drug reaction in 1994, and 106,000 had fatal ADRs. Based on this estimate, adverse drug reactions would be the fourth leading cause of death in the United States.
The single most frequent severe adverse drug reaction is gastrointestinal complications caused by nonsteroidal anti-inflammatory drugs (NSAIDs). The U. S. Food and Drug Administration has estimated that the incidence of symptomatic ulcers and potentially life-threatening ulcer complications in patients taking NSAIDs is 2-4% per year. More than 100 million prescriptions are filled for NSAIDs each year, and many more patients are taking nonprescription products, resulting in over 200,000 hospitalizations and 10,000 to 20,000 deaths due to NSAID-induced GI complications each year (Silverstein, 1998). Considering the frequent use of NSAIDs in arthritis, it is no wonder that NSAID gastropathy has been referred to as the second most deadly rheumatic disease.
Recent investigations into this problem include the Misoprostol Ulcer Complications Outcomes and Safety Assessment (MUCOSA) trial, evaluation of the use of high doses of famotidine or omeprazole for prophylaxis and evaluation of eradication of Helicobacter pylori in at-risk patients. Another promising development is COX-2 selective drugs that appear to have therapeutic efficacy in inflammatory conditions with possibly little or no risk for GI complications.
The MUCOSA trial (Silverstein, 1995) was the first randomized controlled study to determine the efficacy of misoprostol in reducing the incidence of serious GI complications (bleeding, perforation, obstruction) rather than just endoscopically identified ulcers. This is significant because there has been only a modest correlation between endoscopic findings and GI hospitalizations. Misoprostol has been shown to reduce endoscopic ulcers by almost 90% but the MUCOSA study showed a relative risk reduction in clinically significant complications of only 40%. Of the 4,406 patients randomized to receive misoprostol, 25 suffered a serious GI event in the six-month trial, compared to 42 events in the 4,433 patients in the placebo group. This difference represents an absolute risk reduction of 0.38% (p .05), meaning that 263 patients would be treated for six months in order to prevent one serious gastrointestinal event.
Misoprostol was not shown to reduce mortality in this study.
More informed decisions regarding the use of NSAIDs and prophylactic therapy with misoprostol require the identification of patients with additional risk factors such as age - 75 years, previous peptic ulcer, previous GI bleeding, major comorbid conditions, high doses of NSAIDs, concomitant steroids or anticoagulation therapy. The Arthritis Rheumatism and Aging Medical Information System (ARAMIS) is currently developing a more precise calculator based on these factors to estimate the risk of GI events so that the benefits and risks of therapy can be more accurately evaluated (Singh, 1998).
Two recent studies have evaluated high-dose famotidine for the prevention of NSAID-induced ulcer (Hudson, 1997; Taha, 1996). Famotidine 40 mg bid significantly reduced the incidence of gastric and duodenal ulcers. However, these studies were criticized for three important reasons: 1) the definition of ulcer, a mucosal break of 3 mm or more in diameter, could represent clinically insignificant events, 2) the correlation with significant GI events is not certain and 3) the incidence of ulcers in the famotidine treated patients was higher than most studies with misoprostol. Numerous previous trials with histamine-2 receptor antagonists (H2RA) have shown benefit in reducing duodenal but not gastric ulcers in patients taking chronic NSAIDs, and an observational trial (Singh, 1996) has implicated H2RA as a risk factor for increased NSAID-induced GI complications. In spite of the recent study results, famotidine should not be considered effective prophylaxis for NSAID-induced GI events.
Omeprazole has been compared to misoprostol (Hawkey, 1998) and to ranitidine (Yeomans, 1998) for treatment and prevention of NSAID-induced ulcers. Omeprazole was as effective as misoprostol in ulcer healing (71% to 76%) at eight weeks, and was more effective at maintaining remission during six months of continued NSAID therapy after ulcer healing, 61% vs 48% (p = 0.001). Compared to ranitidine, omeprazole was more effective in ulcer healing at eight weeks (80% vs 63%, p .001) and more effective at maintaining remission during six months of continued NSAID therapy (72% vs 59%). The primary criticism of these studies is that they were not designed to show that omeprazole is effective in reducing serious NSAID-associated GI complications or hospitalizations. Much larger studies would be needed to show this difference.
There have been conflicting results regarding the benefits of eradication of H. pylori in reducing the incidence of NSAID-induced peptic ulcers. One randomized trial (Chan, 1997) of eradication of H. pylori prior to initiation of eight weeks of naproxen therapy demonstrated a reduction in the incidence of ulcer: 26% in H. pylori-infected patients vs 3% with successful eradication (p = 0.002). Cross-sectional studies on the risk of H. pylori and NSAID-induced ulcers have shown conflicting results (Wilcox, 1997), and previous prospective trials have not shown an effect of H. pylori on the incidence of NSAID-induced ulcers (Bianchi Porro, 1996). In addition, examination of the results from two trials (Yeomans, 1998; Hawkey, 1998) designed to evaluate the effects of acid inhibition or misoprostol on NSAID-induced ulcers, suggested that H. pylori may offer protection from ulcer in this setting. The physiologic argument for this result is that H. pylori infection increases prostaglandin production. It is possible that there are multiple interactions of various risk factors of age, previous ulcer, use of prophylactic therapy, etc., that will eventually be identified as important for decision making for treatment of patients. Certainly the issue of H. pylori infection and the risk for developing NSAID-induced ulcer is not resolved; however, it is difficult to justify not treating H. pylori infection in a patient who has developed an ulcer.
In spite of many years of research into the incidence, risk factors, mechanisms of injury, pathophysiology and other issues, many unanswered questions remain regarding the most effective and cost-effective clinical management strategies for prevention of NSAID-induced GI complications.
For patients at low risk of NSAID-induced ulcers, no prophylaxis would be indicated considering the low level of potential benefits and the significant associated adverse reactions and costs. Patients infected with H. pylori may benefit from eradication therapy. Patients at high risk of serious gastrointestinal complications from NSAIDs should avoid use of these drugs if at all possible. If NSAIDs must be used, the lowest effective dose, for the shortest duration, and in combination with misoprostol should be considered. For at-risk patients unable to tolerate at least 200 mcg bid of misoprostol, a proton pump inhibitor may be considered for prophylaxis.
We need new studies to assess the relationship of NSAID-induced ulcer complications and the H. pylori status of the patient (infected, uninfected, cured). The relative and absolute benefits and risk of preventive and curative therapies may be significantly different based on this status, and current evidence does not allow this difference to be analyzed. Standard doses of H2RA, sucralfate and antacids are not effective in reducing the incidence of NSAID-associated gastric ulcers and should not be used as prophylaxis for GI events. Further data are needed to confirm the clinical safety and efficacy of the COX-2 selective drugs.