The NSAIDs are a heterogeneous group of compounds that exhibit anti-inflammatory, analgesic, and antipyretic properties. NSAIDs that have been approved by the US Food and Drug Administration (FDA) for OTC analgesic use can be separated into three groups: salicylates, represented by aspirin; propionic acid derivatives, including ibuprofen and naproxen sodium; and the para-aminophenols, represented by acetaminophen (Table 1). While ibuprofen and naproxen are both considered traditional NSAIDs, acetaminophen is not. Although acetaminophen has weak cyclooxygenase (COX) inhibition activity, it appears to have little anti-inflammatory activity, especially in the high-peroxide environment of OA-affected joints2.

The inhibition of prostaglandin production by NSAIDs was first demonstrated in 1971; this work ultimately led to the researchers receiving a Nobel Prize in Medicine3. This activity, distinct from that of other analgesics, was speculated to be the action that mediated the gastric side effects commonly observed with the NSAIDs3. The mechanism of action of the NSAIDs is based on the inhibition of the COX isoenzymes, COX-1 and COX-2. Nonselective NSAIDs inhibit both COX-1 and COX-2, whereas COX-2-specific inhibitors have a minimal effect on COX-1. COX-1 is present constitutively in most normal cells and tissues; it stimulates prostaglandin synthesis, regulates platelet aggregation, and modulates vascular homeostasis, the mucosal integrity of the GI tract, and the functioning of the renal system. COX-2, in contrast, is a key mediator of inflammation; it is induced in response to inflammatory stimuli, initiating pain and the inflammatory response4.

The analgesic effects of NSAIDs have been attributed to the inhibition of COX-2, while the GI side effects are thought to be derived from the inhibition of COX-1. COX-2-specific inhibition was anticipated to preserve effective anti-inflammatory activity while causing little or no associated GI toxicity56. However, the inhibition of COX-2 results in reduced production of prostaglandin I₂ by vascular endothelium, which may contribute to an increased risk for thrombosis7. In addition, antiplatelet effects of NSAIDs are thought to be derived primarily from COX-1 inhibition; thus, the diminished antiplatelet effect is due to the imbalance between prostacyclin and thromboxane production57.

In acute settings, the analgesic effects of non-aspirin NSAIDs are comparable to those of starting doses of opioids58. However, unlike opioids, NSAIDs have a ceiling dose above which no additional pain relief occurs, although higher doses do mediate anti-inflammatory effects8. Clinical studies in patients with arthritis have shown that non-aspirin NSAIDs are comparable to aspirin in reducing pain, joint swelling, and duration of morning stiffness, as well as in improving strength and mobility9. In randomized, double-blind, controlled trials using acetaminophen as the comparator, diclofenac, a nonselective NSAID10, and celecoxib, a COX-2-specific NSAID11, demonstrated greater relief of pain and stiffness.

A growing body of evidence has identified inflammation as an important component of the pathogenesis of OA and OA-related pain1213. Inflammatory cytokines and proteases contribute to a varying extent in the process of joint destruction12 and exacerbation of nociception14. Individual differences in both the degree of inflammation and susceptibility to nociception sensitization may explain, at least in part, why some patients with limited disease radiographically have severe pain, while other patients with more severe deterioration of the joint have minimal pain15. For this reason, the use of analgesics without anti-inflammatory properties in the joints, such as acetaminophen, would only provide pain relief without clinically beneficial anti-inflammatory activity1617.

Among the nonselective NSAIDs, both ibuprofen and naproxen are available in OTC and prescription dosing options. The safety and efficacy of OTC dosing of ibuprofen and naproxen have been assessed in large clinical trials18. OTC doses of both ibuprofen and naproxen provided significantly improved pain control over placebo. Similarly, ibuprofen19 and naproxen20 have been shown to be significantly more effective than acetaminophen for the treatment of pain associated with OA. A comparison of ibuprofen with acetaminophen found that pain relief was superior with ibuprofen, both after single dosing (ibuprofen 400 mg, acetaminophen 1000 mg) and after 14 days of treatment (ibuprofen 400 mg, acetaminophen 1000 mg)19. Similarly, pain relief with naproxen versus placebo at dosages of 660 mg/day (440 mg/day in patients ≥65 years old) was greater than that of acetaminophen versus placebo at dosages of 4000 mg/day20.

Patients without contraindicating conditions (pain persisting for >2 weeks, pain with nausea or severe vomiting, weakness in any limb, pain in a red, hot, or swollen joint, suspected fracture, pelvic or abdominal pain, suspected pregnancy, and increased intensity of pain or any change in the character of pain), can self-treat using OTC NSAIDs21 over a period of 10 days or less. Use of these agents for more than 10 days, even at recommended nonprescription doses, requires the supervision of a physician, as long-term self-treatment could mask a more serious condition requiring a physician's intervention22.

Various clinical guidelines have been proposed for the management of patients with OA454647, the most recent of which was published in February 200844. The 1995 American College of Rheumatology (ACR) guidelines for the medical management of OA of the hip recommended acetaminophen as first-line therapy, permitting the use of a nonselective NSAID if pain control with acetaminophen was inadequate. This recommendation was based on the association of nonselective NSAIDs with GI toxicity46. The updated 2000 ACR guidelines recommended COX-2-specific inhibitors as second-line treatment choices after acetaminophen45. This recommendation was based on the improved GI safety of these agents over that of nonselective NSAIDs, especially in high-risk populations. However, rofecoxib, one of the COX-2 specific inhibitors, was voluntary withdrawn from the US market because of an increased risk of serious CV events33. The ACR guidelines also recommended use of a nonselective NSAID instead of acetaminophen, or with acetaminophen as add-on therapy, if pain control was insufficient45. Similarly, the American Pain Society guidelines from 2002 also recommended acetaminophen as a first-line agent47.

The Osteoarthritis Research Society International (OARSI) has recently published (2008) its recommendations for the management of OA of the knee and hip44. These recommendations are based on a comprehensive review of the published literature and the existing guidelines for the treatment of OA. Consistent with existing guidelines, OARSI indicates that acetaminophen (up to 4 g/day) can be used as initial therapy in patients with mild-to-moderate pain from OA of the knee and hip. If response to acetaminophen is inadequate, or if pain or inflammation (or both) is severe, OARSI recommends that alternative pharmacotherapy be considered. In patients with symptomatic OA of the knee or hip, OARSI recommends use of the lowest effective dose of NSAIDs, for the shortest possible period of time. The important point to note in these 2008 guidelines is that acetaminophen is used for the treatment of mild-to-moderate pain, while NSAIDs are recommended in symptomatic patients who have both pain and inflammation. These updated recommendations are based on previously cited data1617 showing that analgesics such as acetaminophen relieve pain but do not provide anti-inflammatory activity within the joint. Patients at increased GI risk should receive either a proton-pump inhibitor or misoprostol concomitantly with their NSAID for gastroprotection.

In light of the increased risk for GI and CV events in patients with OA, a simple stratification grid was recently proposed to guide clinicians in selecting the most appropriate NSAID for OA patients by evaluating their CV and GI risk factors (Table 2)48. The stratification grid directs clinicians to choose monotherapy with a nonselective NSAID for patients who do not have CV risk factors and are at low risk for GI events. For patients without CV risk but with elevated risk for GI complications and who are not taking aspirin, the recommended treatment is monotherapy with a COX-2-specific inhibitor or combination therapy with a nonselective NSAID plus a proton-pump inhibitor to mitigate GI risk4950. Patients who are at increased CV risk and who require aspirin prophylaxis should be treated with a nonselective NSAID that has been proven to not compromise the antiplatelet effects of aspirin. Although aspirin is approved by the FDA for primary and secondary prevention in a number of CV patient populations51, concomitant administration of aspirin with ibuprofen has been shown to decrease the cardioprotective effects of aspirin37. This effect has not been seen in patients treated concomitantly with aspirin and naproxen38. If these patients require gastroprotection, a proton-pump inhibitor is again suggested; COX-2-specific inhibitors should be avoided in these patients2548.