COX-2-Selective Inhibitors in the Treatment of Arthritis

Affecting nearly 43 million Americans, arthritis is one of the most prevalent diseases and major causes of disability in the United States.(1) By the year 2020, it is estimated that more than 18% of adults in America will have some form of arthritis.(2) Rheumatoid arthritis (RA) is a systemic disease marked by inflammatory changes in synovial membranes and articular structures that lead to widespread degeneration of collagen fibers and destruction of bony structures. Osteoarthritis (OA) is believed to be caused by a combination of abnormal biomechanical stresses on the joint and abnormal biochemical and metabolic changes in the chondrocyte and articular cartilage. Unlike RA, when OA inflammation is present, it is usually mild and localized to the affected joint. Nevertheless, proinflammatory cytokines play a pivotal role in the development of OA disease.(3)

The disease process in OA affects the entire joint and can result in inflammatory changes in the synovium similar to those of RA. These manifest as joint stiffness, loss of physical mobility, and occasionally as joint swelling or redness.(3) Synovial inflammation may be present in early stages of OA, but it is more often seen in advanced stages. OA joint pain, however, does not correlate with histologic evidence of joint inflammation.(4)

Most patients with arthritis are treated by primary care physicians. Therapy for OA is largely palliative, aimed at increasing physical function by relieving joint pain and reducing inflammation.(5) Control of systemic inflammation and prevention or slowing of disease progression are additional treatment goals in patients with RA. While no pharmacologic agents have been shown to prevent or delay the progression of structural damage in OA, disease-modifying antirheumatic drugs (DMARDs) appear to have the capacity to alter the clinical course of RA.(6,7)

Because of their analgesic and anti-inflammatory effects, nonsteroidal anti-inflammatory drugs (NSAIDs) are the class of medication most commonly used to treat joint pain and stiffness in patients with OA and RA.(4,8-10) Nonselective NSAIDs inhibit the isozymes of cyclooxygenase (COX), COX-1 and COX-2. Preclinical studies strongly suggest that inhibition of COX-2 is primarily responsible for many of the therapeutic benefits of NSAIDs, while inhibition of COX-1 can lead to toxic effects.(8,11,12) For this reason, the American College of Rheumatology (ACR) recently recommended replacing nonselective NSAID therapy with therapy with a coxib agent, a COX-2-selective inhibitor, when treating a patient with OA at increased risk of developing an NSAIDrelated toxicity.5 Patients with OA or RA at increased risk of developing NSAID-related gastrointestinal (GI) toxicities include those who are older (65 years of age and above), have a history of a prior symptomatic or complicated ulcer, require chronic high-dose NSAID therapy, or take concomitant corticosteroid or anticoagulant therapy. (4,5,13-15)

The introduction of coxibs represents one of the most rapid development programs of a pharmacologic agent in rheumatology. The first two coxibs, celecoxib and rofecoxib, were approved for use in the United States only a few years after COX-2, the inducible form of COX, was first identified16 and its pathogenic role in pain and inflammation proposed. (17) An aggressive program of clinical trials rapidly followed and provided the evidence-based proof of coxib efficacy in managing the signs and symptoms of OA and RA required by the regulatory approval process.

The efficacy of coxibs in the treatment of osteoarthritis

—Celecoxib, rofecoxib, and valdecoxib are approved for the treatment of OA in the United States.

—The efficacy of celecoxib 100 mg twice daily, 200 mg once daily, and 200 mg twice daily in OA is comparable to that of diclofenac 50 mg three times daily and naproxen 500 mg twice daily and significantly superior to placebo.

—The efficacy of rofecoxib 12.5 mg once daily and 25 mg once daily in OA is comparable to ibuprofen 800 mg three times daily, nabumetone 1,500 mg once daily, and diclofenac 50 mg three times daily and significantly superior to placebo.

—In direct comparisons in OA patients, rofecoxib 25 mg when given once daily in the morning was significantly more effective than celecoxib 200 mg once daily or acetaminophen 1,000 mg four times daily.

—The efficacy of valdecoxib 5 mg once daily, 10 mg twice daily, or 10 mg once daily in OA is comparable to naproxen 500 mg twice daily and superior to placebo.

—Etoricoxib 60 mg once daily and 90 mg once daily are significantly more effective than placebo and comparable to naproxen 500 mg twice daily in the treatment of OA.

—COX-189, an experimental coxib, 50 mg, 100 mg, 200 mg twice daily or 400 mg once daily, provides relief of OA symptoms comparable to diclofenac SR 75 mg twice daily and is significantly superior to placebo.

Outcome Measures in Arthritis Clinical Trials

Clinical trials of pharmacologic agents in OA or RA employ several measures of efficacy recommended by Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT), a group endorsed by the International League of Associations of Rheumatology (ILAR) and the World Health Organization (WHO). These outcome measures are designed to detect minimal clinically significant changes in the severity of joint pain or physical disability associated with OA or RA.(18,19)

Many of these instruments, such as the Patient Assessment of Pain, require evaluation by the patient. The sensitivity and reliability of these selfreport measures have been validated by comparative and radiographic studies.(20-23) One commonly utilized self-rating scale is the visual analog scale (VAS), a continuous numerical scale that ranges from 0 mm, indicative of the best outcome (eg, no pain), to 100 mm for the worst outcome (eg, extreme pain). Another scale often employed in quantifying patient or physician global assessment of disease activity is the Likert scale, a 5-point scale in which 0 designates the best outcome and 4 designates the worst outcome. Minimal clinical significance is generally considered a Likert scale change of at least 0.4 units.(24)

Either the VAS or Likert scale can be used to quantify a patient's status following therapeutic intervention. Many recent OA clinical trials employ the Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC).(18,25) The WOMAC OA Index is composed of 24 items in three subscales that evaluate pain (five questions), physical function (17 questions), and stiffness (two questions).(23,26) Minimal clinically significant change is considered a decrease of 9.7, 9.3, and 10 mm, respectively, in the WOMAC pain, physical function, and stiffness subscales (VAS).(24) The Lequesne Algofunctional Index, graded on a composite scale ranging from 0 to 24, with lower scores indicating better condition, is an outcome instrument commonly employed in clinical trials of hip or knee OA conducted in Europe.(26,27)

The common outcome measure used in RA trials is the ACR 20.28 The ACR developed a binary outcome measure of response based on the seven items in the ILAR/WHO core set. These include the number of painful/tender and swollen joints determined by physical examination, the duration of morning stiffness, patient and physician global assessment of disease activity, severity of pain, a measure of physical disability (eg, Health Assessment Questionnaire [HAQ]), and a measure of an acute-phase reactant (eg, the erythrocyte sedimentation rate or C-reactive protein). To achieve an ACR 20 response, the patient must have at least a 20% improvement in the number of painful/tender and swollen joints as well as an improvement of 20% or more in three of the remaining five outcome measures. While originally developed for use in randomized placebo-controlled trials of DMARDs, the ACR 20 is now widely used in trials of NSAIDs, including COX-2-selective inhibitors.

With few exceptions, all clinical trials of coxib efficacy in OA or RA to date were designed to establish efficacy in patients who had been previously treated with an NSAID and who had experienced a "flare" in symptoms after discontinuing NSAID therapy shortly before study enrollment. (For further discussion of the "withdrawal flare" trial design, see Scott-Lennox et al, 200129). When an NSAID was the active comparator, the higher antiinflammatory dose of NSAID was generally employed. Most of these coxib trials were shortterm, conducted for 6 or 12 weeks. The exceptions were two long-term studies, of 52 weeks' duration, in OA patients comparing rofecoxib with diclofenac,11,30 and one 24-week study comparing celecoxib with diclofenac SR in patients with RA.(31) Two studies of the new coxib, etoricoxib, include a 46-week study versus diclofenac in OA patients32 and a 52-week study comparing etoricoxib with naproxen in OA patients.(33) (See Tables 1 and 2 for trial summaries; following page)

Schnitzer TJ, Hochberg MC. COX-2-selective inhibitors in the treatment of arthritis. Cleve Clin J Med 2002;69:SI20-30.