Published Online: March 17, 2010
Joseph Flood, MD, FACR
The major clinical guidelines recommend the use of acetaminophen (acetyl-para-aminophenol [APAP]) for the treatment of mild-to-moderate symptoms of osteoarthritis (OA) and only recommend the use of nonsteroidal anti-inflammatory drugs (NSAIDs) after APAP failure. This recommendation is based on the efficacy of APAP in treating OA and its relatively benign side-effect profile compared with NSAIDs. NSAIDs are associated with a high risk of adverse events, particularly those of the gastrointestinal (GI) tract. A large number of studies in OA have compared APAP with a variety of selective and nonselective NSAIDs and typically found greater efficacy with NSAIDs. This advantage, however, is mainly the result of increased efficacy in patients with more severe disease, and is viewed as a relatively small analgesic advantage in some studies and meta-analyses. Many of these same studies have reported little or no difference in safety between APAP and NSAIDs, but these results are typically based on short-term studies. Results from meta-analyses on the safety of NSAIDs almost unanimously confirm elevated risk of GI complications. The analgesic mechanism of APAP is still not well understood. However, the notion that APAP has no anti-inflammatory effect has been challenged in recent years with increasing data that suggest it may have an effect on inflammation distinct from that seen with NSAIDs. A variety of mechanistic hypotheses have been proposed.
(Am J Manag Care. 2010;16:S48-S54)
There has been an ongoing struggle to understand and to compare acetaminophen (acetyl-para-aminophenol [APAP]) with nonsteroidal anti-inflammatory drugs (NSAIDs), both selective and nonselective, in the symptomatic treatment of osteoarthritis (OA). Within these larger debates are 2 central questions that are often revisited. First, how does APAP compare with NSAIDs for the symptomatic relief of OA? Second, are concerns regarding the safety and tolerability of NSAIDs sufficient to recommend APAP as an initial treatment in OA?
The major clinical guidelines-including those created by the American College of Rheumatology, European League Against Rheumatism (EULAR), Osteoarthritis Research Society International, and the United Kingdom's National Institute for Health and Clinical Excellence-favor APAP, recommending it as the first choice for mild-to-moderate OA-related pain because of its safety and effectiveness.1-5 The EULAR guidelines additionally state that if APAP treatment is successful, it should be used for long-term analgesia.3,4 The guidelines further recommend that if APAP fails, NSAIDs should be given at the lowest effective dose to avoid or reduce side effects.1-5
Complicating the question of the precise roles of APAP and NSAIDs in the treatment of OA is the fact that OA itself is a rather complex disease. An understanding of the mechanism related to the efficacy of symptomatic treatments for OA, as well as a deeper understanding of OA itself, have occurred in the realm of inflammatory processes (among other aspects of OA). That is, the role of inflammation in OA is now better understood and more prominent in the accepted model of the disease process. Previous assumptions about the anti-inflammatory activity of APAP, or lack thereof, have also evolved and been challenged in recent years.
Efficacy and Safety Studies of Acetaminophen and NSAIDs
In a milestone study from 1991, Bradley and colleagues found that in the symptomatic treatment of knee OA, APAP 4000 mg/day was comparable to ibuprofen at both an "analgesic dose" (1200 mg/day) and an "anti-inflammatory dose" (2400 mg/day).6 Since then, numerous studies have compared APAP with various NSAIDs in OA, more often than not for knee OA, and the results generally support that NSAIDs provide greater pain relief than APAP.7-11 Moreover, a substantial portion of these individual studies have observed little difference with regard to safety and tolerability among NSAIDs and APAP.7,8,12 Before describing the inherent problem with these safety results, it is worth noting that several key meta-analyses have come to very different conclusions. For example, a meta-analysis by Zhang et al compared APAP to NSAIDs in OA and found that while efficacy was greater with NSAIDs, gastrointestinal (GI)-related safety issues were also significantly greater with NSAIDs.9 Furthermore, a 2006 Cochrane review of APAP in OA determined that NSAIDs were more effective than APAP for pain relief, but the magnitude of the difference in treatment effect was "small to modest."13 Because the median length of the reviewed studies was only 6 weeks, the safety and tolerability data derived from them for comparative purposes was of limited practical utility since the medications in question are not generally confined to short-term use.13
An illustrative example of the limitations of safety data from short-term trials is the 2 VACT (Vioxx, Acetaminophen, Celecoxib Trial) studies. The cyclooxygenase (COX)-2 inhibitors, rofecoxib and celecoxib, demonstrated greater efficacy in knee OA compared with APAP, but no significant differences in the incidence of adverse events (AEs) were observed between the treatment groups.7 It should be noted that this 6-week study was published approximately 9 months after rofecoxib was withdrawn from the market due to increased risk of cardiovascular events with long-term use.
Several other studies produced results inconsistent with those from the preponderance of APAP versus NSAID studies. For example, in a 12-week study by Case et al that compared APAP, diclofenac sodium, and placebo in knee OA, not only was the NSAID superior to APAP for symptomatic treatment, APAP was no better than placebo.10 However, the study population included only 25 patients in the diclofenac group, 29 in the APAP group, and 28 in the placebo group. The authors stated that the study was sufficiently powered to reach their conclusion about APAP, although they acknowledged that subset analysis-to determine which patients responded better among individual treatments-was not possible due to the limited number of patients in the study.10
The results of this study were similar to an earlier and larger crossover study by Pincus et al, which compared diclofenac plus misoprostol (for the prevention of GI side effects) with APAP in 218 patients with knee OA. That study found that 57% of subjects preferred treatment with diclofenac/misoprostol, 20% preferred APAP, and 20% had no preference.11 Consistent with clinical guidelines recommending APAP in mild and moderate OA, a subgroup analysis of patients in the Pincus study found that the greatest difference in efficacy occurred among patients with more severe OA, and minimal differences were seen in patients with a milder form of the disease.11 It should be noted that significantly more patients given diclofenac and misoprostol experienced AEs, particularly GI AEs, compared with patients receiving APAP.11 For individual GI AEs, significantly more patients in the diclofenac and misoprostol group experienced abdominal pain and abnormal serum glutamic oxaloacetic transaminase levels compared with those given APAP.11 The incidence of dyspepsia was similar among the 2 treatment groups.
Other studies of APAP compared with placebo have generally found APAP to be superior. Pooled results from 2 identical, multicenter, randomized, double-blind, placebo-controlled studies demonstrated that both APAP and naproxen were better than placebo across 7 symptom domains; naproxen was more effective than APAP; and naproxen, APAP, and placebo had similar safety profiles (although the studies were only 7 days in duration)12 (Figure 1). Neither patient nor investigator evaluations, however, were significantly different regarding the efficacy of naproxen compared with APAP, but both treatments were rated significantly better than placebo by both patients and investigators.12 In a 7-day study in knee and hip OA, APAP was preferred nearly 2 to 1 over placebo.14 In 2004, Pincus et al published results from 2 identical, randomized, double-blind, placebo-controlled, crossover clinical trials in patients with hip and knee OA that compared APAP, celecoxib, and placebo.15 Pain, as measured by visual analog scores, was not significantly different between APAP and celecoxib in the first study. APAP and celecoxib were both superior to placebo. In the second study, celecoxib was significantly superior to both APAP and placebo, and APAP was not significantly better than placebo.15 Patients preferred celecoxib, followed by APAP, and then placebo. In a 6-week, double-blind, parallel-group, placebo-controlled trial, APAP was compared with placebo in patients with knee OA. No significant differences in pain intensity were observed between patient groups; however, both APAP and placebo produced significant reductions in pain.16
PDF is available on the last page.