Prevention of NSAID-Induced Ulcers.

Cyclooxygenase-2 (COX-2) selective inhibitors were developed as gastrointestinal (GI) safer alternatives to nonselective NSAIDs--providing equivalent analgesic and anti-inflammatory efficacy. Their GI sparing is impaired by concomitant aspirin use, and concerns regarding adverse cardiovascular effects have emerged. Risk factors for NSAID-related complications include a history of ulcer disease or bleeding, concomitant corticosteroid or anticoagulant therapy, use of high-dose or multiple NSAIDs (including low-dose aspirin), advanced age, and certain chronic diseases. If an NSAID must be used in a patient at risk, the lowest-risk NSAID should be used with, in many cases, cotherapy to reduce the risk for ulcers. COX-2 drugs have been associated with a significantly higher risk of vascular events than placebo or naproxen. This increase may be shared by nonselective NSAIDs and appears to be medication- and dose-dependent. Prostaglandin depletion is a central mechanism for NSAID ulcer development, and replacement therapy with misoprostol reduces NSAID toxicity; however, it is rarely used due to side effects. The acid suppression provided by traditional doses of histamine 2-receptor antagonists (H(2)RAs) does not prevent most NSAID-related gastric ulcers. Despite a single study demonstrating that H(2)RAs at double the dose may be effective, studies comparing such high doses with misoprostol or proton pump inhibitors (PPIs) for preventing NSAID ulcers are not available. PPIs are effective at single daily doses, do not demonstrate tachyphylaxis, and are superior to H(2)RAs and misoprostol in reducing ulcers and NSAID-associated dyspepsia. NSAID choice should be predicated by an assessment of an individual's cardiovascular and GI risk. For those with competing cardiovascular and GI risks, the tradeoffs between reducing adverse GI events (COX-2 inhibitor instead of a nonselective NSAID) must be explicitly weighed against concerns about cardiovascular side effects (naproxen instead of other agents). Considering cost is appropriate because it may not be feasible to recommend the "safest" regimen in every circumstance. The cost effectiveness of risk-reducing therapies is intimately tied to the patient's underlying risk. For those at highest GI risk, using a PPI and a low-dose COX-2 inhibitor seems appropriate for those without high cardiovascular risk. For patients whose cardiovascular risk parallels or exceeds GI concerns, naproxen with a PPI is recommended when non-NSAID approaches fail.