COX-1-sparing NSAIDs--is the enthusiasm justified?

Abstract

Listen

Translate article

NONSTEROIDAL ANTI-INFLAMMATORY DRUGS (NSAIDS) are among the most widely used therapeutic agents today, with nearly $2 billion spent in the United States yearly on prescription NSAIDs alone. By inhibiting the enzyme cyclooxygenase (COX) and decreasing prostaglandin synthesis, these agents provide effective analgesia and suppress inflammation. Most NSAIDs (nonselective) inhibit not only prostaglandins at sites of inflammation but also prostaglandins that serve important functions in other parts of the body. This inhibition may be beneficial when, for example, these drugs are prescribed to impair normal platelet function to prevent cardiovascular events. However, such inhibition may be harmful when normal gastrointestinal mucosal function is impaired and unwanted mucosal damage occurs. While such damage does not usually lead to symptoms, gastrointestinal ulceration may produce pain and, more ominously, lead to bleeding, perforation, or obstruction. A large prospective trial of more than 8000 patients with rheumatoid arthritis who were taking NSAIDs found the rate of serious gastrointestinal complications to be 0.74% over a 6-month period, or approximately 1.5% per year. Although not a high absolute rate, thousands of patients are at risk for serious adverse effects each year because of the large number of NSAID users. The rates of gastrointestinal complications are even higher among certain subgroups, such as older patients with prior ulcers. Efforts to prevent these NSAID-induced complications have included the concomitant administration of acidsuppressant drugs or a prostaglandin analog. A new approach to avoiding the problem of NSAID complications became feasible with the discovery that there were 2 isoforms of COX, COX-1 and COX-2. COX-1 is expressed normally in the gastrointestinal tract, the kidneys, and platelets and plays major roles in their normal function. COX-2 is expressed primarily in response to inflammation throughout the body and to some degree normally in the kidneys and other body sites. Traditional NSAIDs inhibit both COX-1 and COX-2, providing benefits at sites of inflammation but at the cost of potential adverse effects in areas such as the gastrointestinal tract. Development of NSAIDs that preferentially inhibit COX-2 offered the promise of relieving pain and inflammation without the adverse effects attendant to COX-1 blockade. The US Food and Drug Administration has approved 2 new NSAIDs that specifically inhibit COX-2. The first such drug to be marketed, celecoxib, is at present one of the best-selling new prescription drugs. The second drug, rofecoxib, has just been released to much fanfare. Yet despite this initial enthusiasm, it is unclear whether all the hype is warranted. The relative COX-2 selectivity of NSAIDs has been assessed most commonlyusing invitroassays that compare theamount of COX inhibition in a COX-1 exclusive assay system with the extentof inhibition inaCOX-2exclusive system.Mostof these in vitro COX assay systems have used cell culture, purified enzymes, or nonhuman tissues. Consequently, a valid criticism has been that comparative rankings of COX selectivity in these systems may not reflect COX inhibition in the human joint or, more important, lack of COX inhibition in the gastrointestinal tract. Given that the primary limitation of traditional NSAIDs is gastrointestinal tract adverse effects, the most relevant means to determine the COX selectivity of NSAIDs might be to assess their COX inhibition directly in the stomach,anorganthatprimarilyexpressesCOX-1.Since thedesired therapeutic goal of this new class of NSAIDs is prevention of gastric COX-1 inhibition, an alternative classification is to describe them as COX-1–sparing NSAIDs. A recent comparison of gastric COX inhibition in healthy humans indicates that rofecoxib and celecoxib spare gastric COX-1 to a much greater extent thantraditionalNSAIDs(B.CryerandM.Feldman,MD, unpublished data, 1999). However, the issue of whether this gastricCOX-1–sparingeffect translates intopreventionofuntoward gastrointestinal tract consequences for patients who take these drugs is unresolved. Two articles in this issue of THE JOURNAL are among the first published peer-reviewed multicenter studies of COX-1– sparing NSAIDs. The first, by Simon and colleagues, is a 12week, randomized trial of celecoxib, naproxen, and placebo