Recommendations in the ACR guidelines suggested that, after instituting appropriate nonpharmacologic programs, analgesics and anti-inflammatory agents should be considered. The use of acetaminophen in doses up to 4 g per day was recommended as initial therapy. However, in patients with evidence of inflammation or with severe pain, other initial therapeutic approaches, including nonsteroidal anti-inflammatory agents (NSAIDs) or intra-articular steroids, should be considered. Simple analgesics include acetaminophen followed by agents such as tramadol, propoxyphene, and—in appropriate cases—opioids.
Most patients with OA benefit to various degrees when administered NSAIDs, either traditional nonselective NSAIDs or COX-2 selective agents (called coxibs). COX-2 selective agents have a significant advantage over traditional NSAIDs; they cause fewer adverse gastrointestinal events such as peptic ulcers with associated bleeding, perforation, or obstruction. The observation that some coxibs were associated with an increased rate of cardiovascular events (particularly myocardial infarction) raised concern that increased cardiovascular risk might outweigh the gastrointestinal benefit of the coxibs.8 Coxibs are thought to increase CV risk by causing an imbalance between inhibition of thromboxane and prostacyclin. Coxibs would shift the tendency toward thrombosis by inhibiting prostacyclin, a thrombosis-inhibiting prostaglandin, while not inhibiting thrombosis-promoting thromboxane.
The relationship of nonselective NSAIDs, which effectively inhibit platelets and thromboxane, to CV events is less readily explained.9 Increased thrombotic CV events seen with nonselective NSAIDs may occur because effective inhibition of platelet aggregation requires sustained, over-80% inhibition of platelet COX-1, a level achieved by aspirin and high-dose naproxen. A second mechanism important as a potential cause of increased CV events with NSAIDs of all types, however, relates to increases in hypertension and edema seen to various degrees both with nonselective NSAIDs and coxibs.
A recent statement regarding the treatment of OA by the American Heart Association described a stepped care approach to pharmacologic therapy for osteoarthritis.10 The suggestion was made that initial therapy should include acetaminophen, nonacetylated salicylates, and short-term narcotic analgesics. These recommendations are not evidence based in several respects. Firstly, nonacetylated salicylates (like coxibs) are COX-1 sparing and their potential for increased cardiovascular risk has not been defined. Narcotic analgesics, particularly in older individuals, are associated with significant side reactions, including constipation, cognitive impairment, and the risk of falling. Further, the suggestion that one can differentiate clinical responses to coxibs on the basis of their COX-2–COX-1 specificity ratio has not been clinically validated.