Pharmacotherapy A Pathophysiologic Approach, 9th Ed.

71. Osteoarthritis

Lucinda M. Buys and Mary Elizabeth Elliott


 Images Millions of Americans have osteoarthritis (OA). OA prevalence increases with age, with women more commonly affected than men.

 Images Contributors to OA are systemic (age, genetics, hormonal status, obesity, occupational, or recreational activity) and/or local (injury, overloading of joints, muscle weakness, or joint deformity).

 Images OA is primarily a disease of cartilage that reflects a failure of the chondrocyte to maintain proper balance between cartilage formation and destruction. This leads to loss of cartilage in the joint, local inflammation, pathologic changes in underlying bone, and further damage to cartilage triggered by the affected bone.

 Images Manifestations of OA are local, affecting one or a few joints; the knees are most commonly affected, as well as the hips and hands. Osteophytes (bony proliferation of affected joints) are often found, in contrast to the soft tissue swelling of rheumatoid arthritis.

 Images Nonpharmacologic therapy is the foundation of the treatment plan for all patients with OA. Nonpharmacologic therapy should be initiated before or concurrently with pharmacologic therapy.

 Images The most common symptom associated with OA is pain, which leads to decreased function and motion. Pain relief is the primary objective of medication therapy.

 Images Based on efficacy, safety, and cost considerations, scheduled acetaminophen, up to 4 g/day in divided doses, should be tried initially for pain relief in knee and hip OA. If this fails, topical or oral nonsteroidal antiinflammatory drugs (NSAIDs) are recommended, if there are no contraindications.

 Images Strategies to reduce NSAID-induced GI toxicity include the use of nonacetylated salicylates, COX-2 selective inhibitors, or the addition of misoprostol or a proton-pump inhibitor. COX-2 selective inhibitors vary in their ability to prevent GI toxicity, and concomitant use of aspirin largely negates their gastroprotective effects.

 Images COX-2 selective inhibitors can increase the risk of cardiovascular events. This may be a class effect, but the extent of this risk varies among COX-2 selective inhibitors, and traditional NSAIDs may also pose risks. This hazard, in addition to the GI toxicity possible with all NSAIDs, underscores the importance of using NSAIDs only as needed and after assessing the individual patient’s risk.

 Images NSAIDs are associated with GI, renal, cardiovascular, liver, and CNS toxicity. Monitoring with complete blood count, serum creatinine, and hepatic transaminase levels can be valuable in detecting potential toxicity.

 Images Topical NSAIDs are recommended for patients older than 75 years to decrease the risks of systemic toxicity.

 Images Other agents useful in treating knee OA include tramadol, intraarticular injections of corticosteroids, and duloxetine.

Osteoarthritis (OA) is the most common joint disease and is one of the leading causes of disability in the United States.13 Knee OA alone is as important a contributor to disability as cardiovascular disease and more important than other comorbidities in this respect. In 2009, OA was ranked fourth as a cause for hospitalization in the United States.3

The progressive destruction of articular cartilage has long been appreciated in OA, but OA involves the entire diarthrodial joint, including articular cartilage, synovium, capsule, and subchondral bone, with surrounding ligaments and muscles also playing important roles. Changes in structure and function of these tissues produce clinical OA, characterized by joint pain and tenderness, with decreased range of motion, weakness, joint instability, and disability.

This chapter will review the epidemiology, etiology, pathogenesis, and diagnosis of OA. It then will focus on nonpharmacologic and pharmacologic treatments for OA, as well as investigational agents. Because millions of persons take medications for OA, the overall risks posed by these medications require serious consideration, particularly by clinicians who treat or advise patients on drug therapy for OA. This chapter examines the risks and benefits of OA treatments, with emphasis on those individuals who have the highest risk for adverse events, to help clinicians maximize benefits and reduce risks to their patients with OA.


Images Twenty-seven million adults in the United States were estimated to have clinical OA in 2005, which represented an increase from 21 million in 1995.4 Prevalence of arthritis-related disability is expected to increase to 11.6 million in the United States by 2020, and an estimated 67 million persons will have OA by 2030.3,5

OA imposes a tremendous cost burden, with total medical costs for OA approximately $336 billion in 2004.6 For 2005, out-of-pocket and insurer expenditures were $36 billion and $149 billion, respectively.7In 2009, more than 900,000 knee and hip replacements were carried out, at a cost of $42 billion, with most of these surgeries necessitated by OA.3

Prevalence by Age, Sex, and Race

Prevalence estimates for OA vary depending on the age group of interest, gender, ethnic group, and the specific joint involved. Estimates also depend on the specific means by which OA is assessed and documented. Clinical OA is based on physical exam and patient history, whereas radiographic OA is determined by X-ray or other imaging, and symptomatic OA is based on history and physical plus X-ray.1,3,5

OA is more prevalent with increasing age. In the United States, prevalence of radiographic knee OA was estimated as 13.8% for all persons over age 25, but 37.4% for persons age 60 and older.4 Prevalence for symptomatic knee OA was estimated as 4.9% for all persons over age 25, but 12.1% for persons age 60 and older.4 Radiologically confirmed hip OA shows clear trends through all age groups, affecting 1.6% of those between ages 30 and 39, up to a prevalence of 14% in those over 85 years of age.8 Radiographic hand OA is found in 5% of those age 40, but in 65% of those over 80 years of age.9

Prevalence of hip OA is 9% in white populations, but is only 4% for Asian, black, and Indian populations. Before age 50, men are more likely to have OA than women, attributed to higher rates of sports and other injuries.10Women exhibit a higher prevalence of hip and knee OA than men, and are at especially greater risk for hand OA, with 26% of women and 12% of men over age 70 affected.9 Women are also more likely to have inflammatory OA of the proximal and distal interphalangeal joints of the hands, giving rise to the formation of Bouchard’s and Heberden’s nodes, respectively (Fig. 71-1).


FIGURE 71-1 Heberden’s nodes (distal interphalangeal joint) noted on all fingers and Bouchard’s nodes (proximal interphalangeal joint) noted on most fingers. (From Johnson BE. Arthritis: Osteoarthritis, Gout and Rheumatoid Arthritis. In: South-Paul JE, Matheny SC, Lewis EL, eds. Current Diagnosis and Treatment in Family Medicine. New York: McGraw-Hill, 2004:266, 267.)


The incidence of symptomatic OA determined in a large managed care organization was 100 per 100,000 patient-years for hand OA, 88 per 100,000 patient-years for hip OA, and 240 per 100,000 patient-years for knee OA.8 Using a population database of approximately 4 million persons, a recent Canadian study estimated the annual incidence rate of physician-diagnosed OA in men to be 11.6 per 1000 in the period 2003 to 2004, similar to the rate of 11.3 per 1000 estimated for 1996 to 1997.11 For women, the incidence rate significantly increased from 14.7 to 16.7 per 1000 from the period 1996 to 1997 to the period 2003 to 2004. Some of the increase observed for women could have resulted from the aging of the population and women’s longer life expectancy.


Images The etiology of OA is multifactorial and complex, with development of OA depending on interplay between factors such as genetic predisposition and joint injury.1 Many patients have more than one risk factor for the development of OA. The most common risk factors for the development of OA include age, obesity, gender, occupation, participation in certain sports, history of joint injury or surgery, and genetic predisposition.


Obesity is the most important preventable risk factor for OA. This linkage is strongest for knee OA, although hip OA may also be linked with weight. Hand OA does not appear to be linked. As the epidemic of obesity spreads in the United States and in other developed countries, so will the burdens imposed by OA.5 Obesity often precedes OA and contributes to its development, rather than occurring as a result of inactivity from joint pain.12 In an 11-year study of approximately 30,000 Norwegian men and women, obesity significantly increased the risk of developing OA.13 Men who were obese at baseline had a 2.8-fold increased risk of developing knee OA compared with nonobese men, whereas women who were obese at baseline had a 4.4-fold increased risk of developing knee OA compared with nonobese women. Also, there was an increased risk for severe knee OA in obese subjects.

In addition to being a risk factor for OA, obesity is also a predictor for eventual prosthetic joint replacement. In a US study, women who were obese at age 18 were at increased risk of undergoing hip replacement surgery in later life.3 The risk of developing OA increases by approximately 10% with each additional kilogram of weight, and in obese persons without OA, weight loss of even 5 kg (11 pounds) decreases the risk of future knee OA by half.12

Occupation, Sports, and Trauma

OA risk is increased for people in occupations involving excessive mechanical stress. Work that involves prolonged standing, kneeling, squatting, lifting, or moving of heavy objects increases risk of OA. Such occupations include construction, mining, health care assistance, factory work, carpentry, and farming.1,3,14 Repetitive motion also contributes to hand OA, with the dominant hand usually affected.9 Risk for OA depends on the type and intensity of physical activity and whether injury is incurred in the activity. Increased risk of OA is associated with participation in activities such as wrestling, boxing, baseball pitching, cycling, and football, although recreational participants do not have the increased risk seen in the professional athlete.1,3 In the above-referenced study of 30,000 Norwegians, exercise intensity was not associated with any increased risk in the obese subjects compared with those of normal weight.13

Traumatic injury to articular cartilage during sports and other activities or in accidents greatly increases OA risk.1,3,14 Meniscal damage increases the risk of knee OA because of the loss of proper load bearing and shock absorption, increased focal load on cartilage and on subchondral bone. Knee injury in young persons is also an important risk factor for knee OA in old age.1 Quadriceps muscle weakness is also recognized to increase the risk for knee OA, as these muscles are important in maintaining joint stability.10,12 Whether knee malalignment increases risk of developing OA remains unsettled.1 In the person who already has OA, knee malalignment is strongly associated with faster progression of OA.1

Genetic Factors

OA is a complex, polygenic disease. Identification of the genes involved may promote development of agents to prevent OA or to slow or halt its progression.

Genetic influences on OA have been appreciated for many years. Heberden’s nodes are 10 times more prevalent in women than in men, for example, with a twofold higher risk if the woman’s mother had them. Genetic links have been shown with OA of the first metatarsophalangeal joint and with generalized OA. Twin studies indicate that OA can be attributed substantially to genetic factors (39% to 65%, 60%, and 70% for hand, hip, and spine OA, respectively).15 In other twin studies of OA progression, radiographic measurements over 2 years showed that the increased risk for a sibling having radiographic progression if the proband had progression that was threefold for joint space narrowing and 1.5-fold for osteophyte progression.16

One approach OA researchers have used is the candidate gene approach. This hypothesis is based and focuses on genes with known function that could be plausibly linked with the disease. Genome-wide association studies (GWAS) (associating OA with a specific region out of the total human genome, using cases versus controls, offers a powerful approach in seeking the genetic basis for OA.17,18 Using GWAS and candidate gene approaches, possible genetic associations to OA have been found, and some of these appear to code for known proteins that have intriguing connections.1820 These genes include Col11A1 (extracellular matrix), Chrom 19 (cartilage morphogenesis), MCFL (pain perception), CHST11 (cartilage morphogenesis), GDF5 (TGF-α signaling), and Chrom7Q22. A meta-analysis of GWAS with 6,709 knee OA cases and 44,439 controls revealed that the Chrom7Q22 locus was very highly significantly associated with knee OA. The locus also included six genes that code for proteins known to be expressed in joint tissues.19

For most genes that appear to be linked to OA, the associations have been weak or modest, even if replicable.17 It is quite likely that the genetic risk of developing OA, like many other diseases, may be substantially determined by a combination of modest genetic differences, and this underscores the point that understanding of the genetics and pathology of OA is in its infancy.


OA falls into two major etiologic classes. Primary (idiopathicOA, the more common type, has no identifiable cause. Secondary OA is that associated with a known cause such as rheumatoid or another inflammatory arthritis, trauma, metabolic or endocrine disorders, and congenital factors.10

The old view of OA as a “wear-and-tear” or degenerative disease, largely focused on joint cartilage, has long been superseded by an appreciation of the dynamic nature of OA and that it represents a failure of the joint and surrounding tissues.21 Some changes in the OA joint may reflect compensatory processes to maintain function in the face of ongoing joint destruction. Not only biomechanical forces but also inflammatory, biochemical, and immunologic factors are involved. An appreciation of the biology and function of normal cartilage can aid in understanding osteoarthritic cartilage and is summarized below.

Normal Cartilage

Function, Structure, and Composition of Cartilage

Articular cartilage possesses viscoelastic properties that provide lubrication with motion, shock absorbency during rapid movements, and load support. In synovial joints, articular cartilage is found between the synovial cavity on one side and a narrow layer of calcified tissue overlying subchondral bone on the other side (Fig. 71-2).22 The layer of cartilage is narrow, with human medial femoral articular cartilage being approximately 2 to 3 mm thick. Despite this, healthy articular cartilage in weight-bearing joints withstands millions of cycles of loading and unloading each year. Cartilage is easily compressed, losing up to 40% of its original height when a load is applied. Compression increases the area of contact and disperses force more evenly to underlying bone, tendons, ligaments, and muscles. In addition, cartilage is almost frictionless, and together with its compressibility, this enables smooth movement in the joint, distributes load across joint tissues to prevent damage, and stabilizes the joint.


FIGURE 71-2 Characteristics of osteoarthritis in the diarthrodial joint. (Courtesy of Dr. D. Gotlieb.)

Strength, a low coefficient of friction, and compressibility of cartilage derive from its unique structure. Cartilage is a complex, hydrophilic, extracellular matrix (ECM). It is approximately 75% to 85% water and contains 2% to 5% chondrocytes collagen and other proteins, proteoglycans, and long hyaluronic acid (HA) molecules.22 The two major structural components in articular cartilage are type II collagen and aggrecans.23 Type II collagen has a tightly woven triple helical structure, which provides the tensile strength of cartilage. Aggrecan is a proteoglycan linked with HA, providing the long aggrecan molecules a high negative charge. These are squeezed together by surrounding fibrils of type II collagen. The strong electrostatic repulsion of proteoglycans held in close proximity gives cartilage the ability to withstand further compression. Within the cartilage ECM are the chondrocytes, the only cells in cartilage, responsible for laying down all the components of cartilage.

Normal cartilage turnover helps repair and restore cartilage in response to demands of joint loading and during physical activity. In adults, cartilage chondrocyte metabolism is slow and is regulated by growth factors, including bone morphogenetic protein 2, insulin-like growth factor-1, and transforming growth factor, and by catabolism and proteolysis stimulated by matrix metalloproteinases (MMPs), tumor necrosis factor-α (TNF-α), interleukin-1, and other cytokines. Tissue inhibitors of metalloproteinase (TIMP) also contribute to the balance by restraining the catabolic actions of MMPs. If cartilage is injured, chondrocytes react by removing the damaged areas and increasing synthesis of matrix constituents to repair and restore cartilage.23,24

Another component supporting healthy joints are the joint protective mechanisms, such as muscles bridging the joint, sensory receptors in feedback loops to regulate muscle and tendon function, supporting ligaments, and subchondral bone that has shock-absorbent properties.

Finally, it is important to note that adult articular cartilage is avascular, with chondrocytes nourished by synovial fluid. With movement and cyclic loading and unloading of joints, nutrients flow into the cartilage, whereas immobilization reduces nutrient supply. This is one of the reasons that normal physical activity is beneficial for joint health.

Osteoarthritic Cartilage

Images Important contributors to the development of OA are local mechanical influences, genetic factors, inflammation, and aberrant chondrocyte function leading to loss of articular cartilage.23,24 At a molecular level, OA pathophysiology involves the interplay of dozens, if not hundreds, of extracellular and intracellular molecules with roles including chondrocyte regulation, phenotypic changes, proteolytic degradation of cartilage components, and interactions between articular cartilage, underlying subchondral bone, and the joint synovium.2,5,2327

OA most commonly begins with damage to articular cartilage, through trauma or other injury, excess joint loading from obesity or other reasons, or instability or injury of the joint that causes abnormal loading. In response to cartilage damage, chondrocyte activity increases in an attempt to remove and repair the damage. Depending on the degree of damage, the balance between breakdown and resynthesis of cartilage can be lost, and a vicious cycle of increasing breakdown can lead to further cartilage loss and apoptosis of chondrocytes.2,5,2327 Recent studies have revealed several respects of the very complex nature of OA. For example, expression of hundreds of specific genes are affected by acute experimental injury of human cartilage tissue, that is, injury alters the chondrocyte phenotype.28 Researchers have also shown that within different regions of human OA cartilage obtained at surgery, chondrocyte gene expression from the most damaged areas of cartilage is different from that from less damaged or normal areas.29 Another exciting discovery is that comparative proteomics of articular cartilage from normal persons compared with cartilage from those with OA showed different expression.30

There is increased appreciation of the role of tissues beyond cartilage, within the joint and surrounding it, subchondral bone.24 Subchondral bone undergoes pathologic changes that may precede, coincide with, or follow damage to the articular cartilage. In OA, subchondral bone releases vasoactive peptides and MMPs, and damage to subchondral bone may trigger further damage to articular cartilage.31Neovascularization and subsequent increased permeability of the adjacent cartilage occur and contributes further to cartilage loss.

Joint space narrowing results from loss of cartilage, which can lead to a painful, deformed joint (Fig. 71-3). Remaining cartilage softens and develops fibrillations (vertical clefts into the cartilage), followed by splitting off of more cartilage and exposure of underlying bone.32 During this time, adjacent subchondral bone undergoes further pathologic changes; cartilage is eroded completely, leaving denuded subchondral bone, which becomes dense, smooth, and glistening (eburnation). A more brittle, stiffer bone results, with decreased weight-bearing ability and development of sclerosis and microfractures. New bone formations, or osteophytes, also appear at joint margins distant from cartilage destruction and are thought to arise from local and humoral factors. There is direct evidence that osteophytes can help stabilize osteoarthritic joints.33


FIGURE 71-3 Plain x-ray films of the knee demonstrating joint space narrowing. (From Johnson BE. Arthritis: Osteoarthritis, Gout and Rheumatoid Arthritis. In: South-Paul JE, Matheny SC, Lewis EL, eds. Current Diagnosis and Treatment in Family Medicine. New York: McGraw-Hill, 2004:267.)

In the joint capsule and synovium, inflammatory changes and pathologic changes can occur.2,22,2427 Contributors to inflammation may include crystals or cartilage shards in synovial fluid. Other possible players are interleukin-1, prostaglandin E2, TNF-α, and nitric oxide, which are found in synovial fluid. With inflammatory changes in the synovium, effusions and synovial thickening occur.

The pain of OA is not related to the destruction of cartilage but arises from the activation of nociceptive nerve endings within the joint by mechanical and chemical irritants.5 OA pain may result from distension of the synovial capsule by increased joint fluid, microfracture, periosteal irritation, or damage to ligaments, synovium, or the meniscus.



Images The diagnosis of OA is made through history, physical examination, characteristic radiographic findings, and laboratory testing.10,34,35 The major diagnostic goals are (a) to discriminate between primary and secondary OA and (b) to clarify the joints involved, severity of joint involvement, and response to prior therapies, providing a basis for a treatment plan. The American College of Rheumatology has published traditional diagnostic criteria and “decision trees” for OA diagnosis.35 As with all guidelines, the authors stress these are for assisting the clinician rather than replacing clinical judgment. For example, traditional criteria are as follows: (a) For hip OA, a patient must have pain in the hip and at least two of the following three: an erythrocyte sedimentation rate <20 mm/h (>5.6 μm/s), femoral or acetabular osteophytes on radiography, or joint space narrowing on radiography. This provides a sensitivity of 89% and a specificity of 91%. (b) For a clinical diagnosis of knee OA, a patient must have pain at the knee and osteophytes on radiography plus one of the following: age older than 50 years, morning stiffness no more than 30 minutes, crepitus on motion, bony enlargement, bony tenderness, or palpable warmth. This provides a sensitivity of 95% and a specificity of 69%. The addition of laboratory or radiographic data further improves accuracy of diagnosis. Criteria for hand OA have also been published.36



    • Usually elderly


    • Age <45 more common in men

    • Age >45 more common in women


    • Pain

    • Deep, aching character

    • Pain on motion

    • Stiffness in affected joints

    • Resolves with motion, recurs with rest (“gelling phenomenon”)

    • Usually duration <30 minutes

    • Often related to weather

    • Limited joint motion

    • May result in limitations of activities of daily living

    • Instability of weight-bearing joints

Signs, History, and Physical Examination

    • Monoarticular or oligoarticular, asymmetrical involvement

    • Hands

      • Distal interphalangeal joints

       • Heberden’s nodes (osteophytes or bony enlargements) (Fig. 71-1)

      • Proximal interphalangeal joints

       • Bouchard’s nodes (osteophytes)

      • First metacarpal joint

       • Osteophytes give characteristic square appearance to hands

      • Knee

       • Pain related to climbing stairs

       • Transient joint effusion

       • Genu varum (“bow-legged”)

      • Hips

       • Groin pain during weight-bearing exercises

       • Stiffness, especially after activity

       • Limited joint movement

    • Spine

       • Lumbar involvement is most common at L3 and L4

       • Paraesthesias

       • Loss of reflexes

      • Feet

       • Typically involves the first metatarsophalangeal joint

       • Shoulder, elbow, acromioclavicular, sternoclavicular, and temporomandibular joints may also be affected

      • Observation on joint examination

       • Bony proliferation or occasional synovitis

       • Local tenderness

       • Crepitus

       • Limited motion with passive/active movement

       • Deformity

      • Radiologic evaluation

       • Early mild OA

       • Radiographic changes often absent

       • Progressive OA

       • Joint space narrowing (Fig. 71-3)

       • Subchondral bone sclerosis

       • Marginal osteophytes

      • Late OA

       • Abnormal alignment of joints

       • Effusions


The prognosis for patients with primary OA is variable and depends on the joint involved. If a weight-bearing joint or the spine is involved, considerable morbidity and disability are possible. In the case of secondary OA, the prognosis depends on the underlying cause. Treatment of OA may relieve pain or improve function but does not reverse preexisting damage to the joint.


Desired Outcome

Management of the patient with OA begins with a diagnosis based on a careful history, physical examination, radiographic findings, and an assessment of the extent of joint involvement. Treatment should be tailored to each individual. Goals are (a) to educate the patient, family members, and caregivers; (b) to relieve pain and stiffness; (c) to maintain or improve joint mobility; (d) to limit functional impairment; and (e) to maintain or improve quality of life.3739

General Approach to Treatment

Treatment for each OA patient depends on the distribution and severity of joint involvement, comorbid disease states, concomitant medications, and allergies. Management for all individuals with OA should begin with both oral and written patient education, a customized activity and exercise program, and weight loss, if the patient is overweight or obese.3739

The primary objective of medication is to alleviate pain.3739 Scheduled acetaminophen, up to 4 g/day in divided doses, should be tried initially (knee, hip), if contraindications are not present. Application of topical NSAIDs over specific joints (knee, hands) and topical capsaicin (hands) are recommended as initial therapy. Nonsteroidal antiinflammatory drugs (NSAIDs) or possibly a cyclooxygenase-2 (COX-2)–selective inhibitor (celecoxib) can be prescribed after careful risk assessment if additional pain control is needed. Intraarticular corticosteroid injections (knee or hip) can relieve pain and are offered concomitantly with oral analgesics or after failed trials of first-line medications, depending on the practitioner’s preference. With centrally acting serotonin reuptake inhibition and analgesic properties, tramadol can also be considered if acetaminophen or topical treatment is ineffective or not tolerated.

Opioid analgesics may be considered if first-line medications are ineffective or pose significant safety concerns in an individual patient. Consideration can also be given to duloxetine or less frequently, HA injections when additional pain control is needed for knee OA. When symptoms are intractable or there is significant loss of function, joint replacement can be appropriate if the patient is a surgical candidate.

There is general agreement that glucosamine and/or chondroitin and topical rubefacients lack uniform efficacy in the treatment of hip and knee OA pain and are not preferred treatment options.

Nonpharmacologic Therapy

Images Nonpharmacologic therapy is an integral part of the treatment plan for all patients with OA.3739 Nonpharmacologic therapy is the only available treatment that has been shown to delay the progression of OA.40 Delaying the progression of OA through active use of nonpharmacologic therapy is critical to prevent future functional impairment. Patient-specific characteristics such as the number and location of affected joints, degree of functional impairment, body mass index, motivation, and overall health status determine which nonpharmacologic therapies should be offered. Nonpharmacologic therapy should be ongoing treatment for all patients, even those who require pharmacologic therapy for pain control (Table 71-1).

TABLE 71-1 Nonpharmacologic Interventions in the Treatment of OA 3739


Patient Education

The first step in OA treatment is patient education about the disease process, the extent of OA, the prognosis, and treatment options. Education is paramount in that OA is often seen as a wear-and-tear disease, an inevitable consequence of aging for which nothing helps. Even worse, patients may resort to the use of alternative but unproven medications or quackery. Organizations such as the Arthritis Foundation provide a wealth of educational information for patients regarding OA, OA medications, local clinics, and agencies offering physical and economic assistance. Exercise, weight loss, and nutritional information are also available. Most educational information is readily available online for patient use.

The benefits of patient education have been documented in a variety of programs.41 These programs are provided across a wide spectrum of delivery methods: from trained volunteers using telephone calls to group sessions for patient support to one-on-one educational sessions with physical therapists or nurse educators. While nearly all of these delivery methods are effective, cost of delivery is highly variable. Long-term cost-effectiveness is very important for sustainability of these patient education programs.

Weight Loss

Excess weight increases the biomechanical load on weight-bearing joints and is the single best predictor of need for eventual joint replacement.42,43 Weight loss of amounts as small as 4% of body weight can lessen OA pain in the knee.44 Greater amounts of weight loss, especially when associated with regular exercise, improve joint function and substantially lessen pain.42,45 At least one randomized controlled trial has demonstrated improvement in pain and self-reported physical function using a combination of modest weight loss (5%) and exercise.46 Patients with appropriate indications for bariatric surgery have significant improvement in joint function and pain associated with the subsequent weight loss.47 A large weight-loss and activity trial is under way, and the Intensive Diet and Exercise for Arthritis trial (IDEA) will address weight-related joint loading and inflammatory biomarkers.48 Weight loss requires a motivated patient, but it should be encouraged and supported for all obese and overweight patients with OA. Effective behavior change strategies should be employed to promote weight loss in patients with OA.40


Exercise programs can improve joint function and can decrease disability, pain, and analgesic use by OA patients.40,44,46 Isometric exercise is preferred over isotonic exercise because the latter can aggravate affected joints. Exercises can be taught and then observed before the patient exercises at home. The frequency, types of exercise, and setting of exercise are still uncertain, but patients who exercise at least two to three times per week with a variety of exercises (>8 types) have improved outcomes.49 The patient should be instructed to decrease the number of repetitions if severe pain develops with exercise.

Some regular exercise should be encouraged for all patients with OA.39 With weak or deconditioned muscles, the load is transmitted excessively to the joints; weight-bearing activities can exacerbate symptoms. Many patients fear that exercise will promote further joint damage and avoid exercise as a means to protect the joint. However, avoidance of regular exercise by those with hip or knee OA leads to further deconditioning and/or weight gain. This leads to more pain and impaired joint function, promoting a downward spiral of disability. Current research regarding exercise revolves around effective strategies to promote sustained behavior change in patients.40 A program of patient education, muscle stretching and strengthening, and supervised walking can improve physical function and decrease pain for patients with knee OA.46

Referral to the physical and/or occupational therapist is especially helpful for developing a customized exercise plan for patients with functional disabilities. The therapist can assess muscle strength and joint stability and recommend exercises and assistive and orthotic devices, such as canes, walkers, braces, heel cups, splints, or insoles for use during exercise or daily activities. Heat or cold treatments help to maintain and restore joint range of motion and to reduce pain and muscle spasms. Warm baths or warm water soaks may decrease pain and stiffness. Heating pads should be used with caution, especially in the elderly. Patients should be warned not to fall asleep on the heat source or to lie on it for more than brief periods to avoid burns.


Surgery can be recommended for OA patients with functional disability and/or severe pain unresponsive to conservative therapy. Criteria for total joint replacement (arthroplasty) of the knee and hip have been developed, although there is substantial overlap in eligibility criteria.50,51

Few randomized, controlled trials are available comparing total joint arthroplasty with other treatment modalities. Although total knee arthroplasty can decrease pain and improve function for many patients, about 20% experience little or no improvement in pain, disability, and/or quality of life.52 These findings coupled with overlapping indications for the procedure and the expected increase in the number of patients with OA lend some urgency to the need for controlled trials to evaluate the outcome of joint replacement with other treatment modalities.

The MEDIC-study of total knee replacement plus physical and medical therapy or treatment with physical and medical therapy alone is currently enrolling patients in hopes of answering some of these important questions about the role of surgery in the treatment of knee OA.52

Total joint arthroplasty is responsible for a large portion of the direct medical costs associated with OA in the United States. The cost-effectiveness of total knee arthroplasty has been evaluated for a Medicare-age population.53Calculations were based on Medicare claims data and costs and outcomes data. Cost projections were calculated for lifetime costs as well as quality-adjusted life expectancy (QALE) for different risk populations and across low-volume to high-volume hospitals. Although total knee arthroplasty was found to be cost-effective across hospital settings and patient risk categories, the procedure was found to be most cost-effective when performed in high-volume centers.

Other surgical options are also available. Arthrodesis (joint fusion) can reduce pain but will restrict motion and may be appropriate for smaller joints that are causing intractable pain. For patients with mild knee OA, an osteotomy (removal of bony tissue) may correct the misalignment of genu varum (“bowlegged” knees) or genu valgum (“knock-knees”). In addition, osteotomies of the pelvis or femur can ameliorate joint misalignment in hip OA, subsequently slowing progression of disease. Knee arthroscopy or lavage appear to be equivalent to sham surgery and are not recommended.37 Experimental but potentially restorative approaches involve soft-tissue grafts, chondrocyte transplantation, gene therapy, and use of growth factors or artificial matrices.54 Cartilage-restoration approaches are investigational, and results regarding pain control and joint function have been mixed.

Pharmacologic Therapy

Images Drug therapy in OA is targeted at relief of pain. OA is commonly seen in older individuals who have other medical conditions, and OA treatment is often long term. As such, a conservative approach to drug treatment, focusing on the needs of the individual patient, is warranted (see Figs. 71-4 and 71-5).3739 Even when pharmacologic therapy is initiated, appropriate nondrug therapies should be continued and reinforced. Specific drug therapy recommendations depend on which joint(s) are affected, response to previous trials of medication, and patient comorbidities.


FIGURE 71-4 Treatment recommendations for knee and hip osteoarthritis.3739 (CV, cardiovascular; NSAID, nonsteroidal antiinflammatory drug.)


FIGURE 71-5 Treatment recommendations for hand osteoarthritis.3739 (NSAID, nonsteroidal antiinflammatory drug.)

Knee and Hip OA

First-Line Treatments

Acetaminophen Images The American College of Rheumatology, as well as others, recommend acetaminophen as a first-line treatment for knee and hip OA.37,39,55 Acetaminophen has been extensively studied in the treatment of knee and hip OA and is more effective than placebo in controlling OA pain.55,56 Compared with oral NSAIDs, acetaminophen may be modestly less effective, but it has a lower risk of serious GI and cardiovascular adverse events and as a consequence is preferred over oral NSAIDs as first-line treatment.55 The significantly lower risks of both minor and major adverse events associated with acetaminophen in the treatment of knee and hip OA favors a trial of acetaminophen in all patients without underlying liver disease.55,56

Oral NSAIDs The American College of Rheumatology and other key groups recommend nonspecific or COX-2 selective NSAIDs, depending on patient risk factors, as a first-line option for knee and hip OA if the patient fails acetaminophen.37,39,55 NSAIDs have a consistent record of providing superior pain relief in comparison to acetaminophen, but no NSAID has proven superior to another.55,56 Nonselective and COX-2–selective NSAIDs pose higher risks for GI, renal, and cardiovascular adverse events compared with acetaminophen. COX-2 inhibitors carry less risk for both minor and serious GI adverse events in comparison to nonselective NSAIDs (with the exception of diclofenac). It is unclear whether the reduced GI risk seen with COX-2 selectivity persists past 3 to 6 months, and this advantage is substantially diminished for patients taking aspirin.55 Proton-pump inhibitors (PPIs) and misoprostol significantly reduce the occurrence of GI adverse events in those taking NSAIDs.55

Topical NSAIDs—Knee Only The American College of Rheumatology and other authoritative organizations recommend topical NSAIDs as a first-line option for knee OA if the patient fails acetaminophen, and is preferred over oral NSAIDs for those older than 75.37,39,55 Randomized trials have demonstrated that topical NSAIDs provide pain relief for OA similarly to that obtained with oral NSAIDs but with fewer GI adverse events. Topical NSAIDs are associated with more frequent local (application site) adverse events compared with oral NSAIDs.55

Intraarticular Corticosteroids Intraarticular corticosteroid injections are recommended as alternative first-line treatment for both knee and hip OA when pain control with acetaminophen or NSAIDs is suboptimal.37,39 Injections can also be administered with concomitant oral analgesic therapy as needed for additional pain control. Intraarticular corticosteroids are generally safe and well tolerated, but should not be administered more frequently than once every 3 months due to risks of systemic adverse effects.

Tramadol Tramadol is recommended as an alternative first-line treatment of knee and hip pain due to OA in patients who have failed treatment with scheduled full-dose acetaminophen and topical NSAIDs, are not appropriate candidates for oral NSAIDs, and are not able to receive intraarticular corticosteroids.39 Tramadol can also safely be added to partially effective acetaminophen or oral NSAID therapy. Fewer data support the use of tramadol as monotherapy for OA pain.

Second-Line Treatments

Opioid Analgesics The American College of Rheumatology recommends opioid analgesics as the primary second-line medication for both knee and hip OA.39 Opioids should be considered in patients who have not had an adequate response to both nonpharmacologic and first-line pharmacologic therapies. Patients who are at high surgical risk, precluding joint arthroplasty, are also candidates for opioid therapy. Opioids provide effective short-term pain control in patients with OA, although data from long-term use trials are less compelling.57 Adverse effects, including serious events, limit the routine use of opioids in the treatment of OA pain. Common adverse events include nausea, vomiting, constipation, somnolence, and dry mouth. Serious events include falls, respiratory depression, and addiction.57

Duloxetine Duloxetine can be used as adjunctive treatment in patients with a partial response to first-line analgesics.37,39 It may be a preferred second-line medication in patients with both neuropathic and musculoskeletal OA pain. Duloxetine has demonstrated efficacy primarily as add-on therapy when there has been less than optimal response to acetaminophen or oral NSAIDs.58,59 Adverse events associated with duloxetine in the treatment of knee and hip OA are most commonly GI with nausea, vomiting, and constipation being the most common. Serious adverse events have not been reported in OA trials that most commonly used moderate doses of 60 mg/day.

Intraarticular Hyaluronic Acid The American College of Rheumatology, the National Institute for Health and Clinical Excellence in the United Kingdom, and others do not routinely recommend the use of intraarticular HA injections for knee OA pain.3739 HA injections do not appear to provide clinically meaningful improvement in pain and/or function scores, although some studies may report statistical differences in scores. These agents may be associated with serious adverse events such as increased pain, joint swelling, and stiffness. Limited efficacy and risks of serious events limit the routine use of these agents.

Hand OA

First-Line Treatments

NSAIDs The American College of Rheumatology and the U.K. National Institute for Health and Clinical Excellence (NICE) recommend topical NSAIDs as a first-line option for hand OA.38,39 Application of diclofenac gel compared with vehicle for hand OA provided significant relief, with mild application-site paresthesia as the only treatment-related adverse effect.60 Topical diclofenac showed similar efficacy as oral ibuprofen as well as oral diclofenac, but with fewer GI adverse events.61,62 Topical diclofenac use was associated with more frequent local (application site) events compared with oral NSAIDs. In all of these studies, topical diclofenac produced fewer GI adverse events.55,61,62

Oral NSAIDs are recommended as an alternative first-line treatment for hand OA by the American College of Rheumatology and as second-line therapy in the NICE guidelines.38,39 For hand OA, there has long been a focus toward topical treatment, perhaps due to reluctance to undergo systemic exposure to strong treatment in patients without pain in a weight-bearing joint.62 Ibuprofen, lumiracoxib, and meclofenamate each provided improvement in hand pain and other symptoms when compared with placebo.61 Active comparator studies for hand OA compared rofecoxib with naproxen and compared lumiracoxib with celecoxib. Efficacy was similar for the comparator NSAIDs in each study, with similar percentages of patients discontinuing due to side effects. For the person who cannot tolerate local skin reactions or who received inadequate relief from topical NSAIDs, oral NSAIDs can offer relief, but the patient then faces increased risk for GI and cardiovascular adverse events.

Topical Capsaicin Capsaicin cream is recommended as an alternative first-line treatment for hand OA.39 Clinical trial data supporting the use of capsaicin for the treatment of hand OA is limited to small studies, but the agent demonstrates modest benefits in improvement of pain scores.61 Adverse effects associated with capsaicin are primarily skin irritation and burning; therefore, it is a reasonable therapeutic alternative for patients not able to take oral NSAIDs.

Tramadol Tramadol is recommended by the American College of Rheumatology as an alternative first-line treatment for OA of the hand.39 In clinical practice, tramadol is a therapeutic option for patients who do not respond to topical therapy and are not candidates for oral NSAID treatment due to high GI, cardiovascular, or renal risks. Tramadol may also be used in combination with partially effective acetaminophen, topical therapy, or oral NSAIDs.


Highlighted drug information will be reviewed below. This section is not intended to be all inclusive, but aims to provide pertinent drug information to facilitate the safe and effective use of these medications in patients with OA.

First-Line Treatments


Pharmacology and Mechanism of Action Acetaminophen is understood to act within the CNS by inhibiting synthesis of prostaglandins, agents that enhance pain sensations. Acetaminophen prevents prostaglandin synthesis by blocking the action of central cyclooxygenase (COX). Acetaminophen is well absorbed after oral administration, with a bioavailability of 60% to 98%. It achieves peak concentrations within 1 to 2 hours, it is inactivated in the liver by conjugation with sulfate or glucuronide, and its metabolites are renally excreted.

Adverse Effects Although acetaminophen is one of the safest analgesics, its use carries some risks, primarily hepatotoxicity and possibly renal toxicity with long-term use.63,64 Serious hepatotoxicity, including fatalities, have been well documented with acetaminophen overdose (see eChap. 10, Clinical Toxicology, for information on treatment of acetaminophen overdose). Continued reports of serious hepatotoxicity, including fatalities from unintentional overdose, have led to labeling revisions of all nonprescription acetaminophen-containing analgesics.65 Unintentional overdoses of acetaminophen are due to a variety of circumstances including narrow therapeutic window at the maximum dose (4 g/day), interpatient differences in sensitivity to liver injury from acetaminophen, a wide array of nonprescription and prescription products that contain acetaminophen and the difficulty of identifying the agent on product labels, and consumers’ lack knowledge about the association of acetaminophen and serious liver injury.

In a study of normal, healthy volunteers administered acetaminophen 4 g/day (1 g every 6 hours), alone or with concomitant opioid therapy, for 14 days, elevations of alanine aminotransferase at levels above three times the upper limits of normal were found in 31% to 44% of patients, depending on the treatment group.66 None of these participants had clinical symptoms of acute liver disease. Although the results of this study are not robust enough to alter the current standard dosing recommendations, it serves as an important reminder that the maximum dose of acetaminophen should be not be exceeded in any patient population and that long-term use of the maximum daily dose of 4 g/day can affect the liver.

Acetaminophen should be used cautiously in patients with liver disease or in those who abuse alcohol. Acute liver failure has been reported in patients taking less than 4 g/daily.67 The most common risk factor for liver failure for these patients was chronic alcohol intake. The FDA has recommended that chronic alcohol users (three or more drinks daily) avoid acetaminophen intake as it increases the risk of liver damage or GI bleeding. Other individuals do not appear to be at increased risk of GI bleeding.

The National Kidney Foundation strongly discourages the use of nonprescription combination analgesic products (e.g., acetaminophen and NSAIDs) because this is associated with an increased prevalence of renal failure.63 A recent large cohort study of patients with chronic kidney disease found acetaminophen, aspirin, and NSAID use to associated with an increased risk of end-stage renal disease (ESRD).64 In addition, the increased risk of ESRD was dose-dependent.

Clinical Controversy…

Is regular use of acetaminophen with an NSAID safe for the kidney?

ESRD caused by phenacetin was recognized more than 20 years ago when the drug was in use, often in combination with an NSAID. The Ad Hoc Review Committee of the International Study Group on Analgesics and Nephropathy studied whether the newer, nonphenacetin-containing combined analgesics were associated with renal disease. The Committee concluded in 2000 that there was not enough evidence to associate nonphenacetin-combined analgesics with nephropathy and that new studies should be done to help resolve the question.

Uncertainty remains about this issue. Most recent OA guidelines do not specifically address the issue of acetaminophen in combination with an NSAID if the patient fails acetaminophen. Some experts recommended that addition of an NSAID to an acetaminophen regimen is reasonable, and some patients take both regularly.

Drug–Drug Interactions and Drug–Food Interactions Drug interactions with acetaminophen can occur; for example, isoniazid can increase the risk of hepatotoxicity. Chronic ingestion of maximal doses of acetaminophen may intensify the anticoagulant effect for patients taking warfarin; such individuals may need closer monitoring.

Although food decreases the maximum serum concentration of acetaminophen by approximately half, the overall efficacy is unchanged.

Dosing and Administration When used for chronic OA, acetaminophen should be administered in a scheduled manner. It may be taken with or without food. Acetaminophen can be taken at 325 to 650 mg every 4 to 6 hours, but the total dose must not exceed 4 g/day (see Adverse Effects above). FDA labeling requirements warn patients about potential liver toxicity if they inadvertently ingest more than the recommended dose when using multiple products containing acetaminophen. Additionally, FDA has also requested that all manufacturers of prescription analgesics containing acetaminophen limit the drug content to 325 mg per tablet or capsule to further decrease the possibility of inadvertent overdoses. Acetaminophen should be avoided in the setting of chronic alcohol intake or in those with underlying liver disease.

Oral Nonsteroidal Antiinflammatory Drugs

Pharmacology and Mechanism of Action NSAIDs reduce pain, inflammation, and fever by preventing synthesis of tissue prostaglandins and related prostanoids, which play a role in triggering these symptoms. All NSAID drugs bind reversibly to the COX-2 enzyme, blocking its action and thus prostanoid production. Blockade of prostaglandin synthesis by inhibiting COX enzymes (mainly COX-2) is thought to account for NSAIDs’ ability to relieve pain and inflammation (Fig. 71-6).68,69 Nonselective NSAIDs were developed before extensive knowledge of COX enzymes was available, but in fact they block both COX-2 and COX-1. COX-1 has required “housekeeping” functions such as gastroprotection. COX-2 inhibitors selectively block COX-2 and have no COX-1 activity.


FIGURE 71-6 Pathway of synthesis for prostaglandins and leukotrienes. COX-1 and COX-2 are cyclooxygenase-1 and cyclooxygenase-2 enzymes, respectively. The minus (–) sign indicates inhibitory influence. Prostaglandins include PGE2 and PGI2; the latter is also known as prostacyclin.

The various NSAIDs exhibit several pharmacokinetic similarities, including high oral availability, high protein binding, and absorption as active drugs (except for sulindac and nabumetone, which require hepatic conversion for activity). There is a broad range of serum half-lives for different NSAIDs, which influence dosing frequency, and potentially, compliance with therapy.70 Elimination of NSAIDs largely depends on hepatic inactivation, with a small fraction of active drug being renally excreted. NSAIDs penetrate joint fluid, reaching approximately 60% of blood levels.

Adverse Effects

Gastrointestinal Effects of Nonselective NSAIDs Images The most common adverse effects of NSAIDs involve the GI tract, contributing to many treatment failures.7173 Minor complaints—nausea, dyspepsia, anorexia, abdominal pain, flatulence, and diarrhea—affect 10% to 60% of patients. All NSAIDs increase ulcer risk, but the serious GI complications associated with NSAIDs include perforations, gastric outlet obstruction, and bleeding. These important GI complications occur in 1.5% to 4% of patients per year. NSAIDs are so widely used that these small percentages translate into substantial morbidity and mortality. Moreover, the risk increases substantially for patients with risk factors including a history of complicated ulcer, concomitant use of multiple NSAIDs (including aspirin) or anticoagulant medications, use of high-dose NSAIDs, and age older than 70 years.74 Consequently, about 16,500 deaths and 103,000 hospitalizations in the United States are associated annually with NSAID use in rheumatoid arthritis or OA patients.

Several options are available for reducing the GI risk of traditional NSAIDs:

   1. Take the lowest dose of the NSAID possible, and take only when needed.

   2. With the NSAID, take the prostaglandin analog misoprostol four times daily. It reduces the rate of ulcers and serious GI complications. However, many patients cannot tolerate the GI adverse events of misoprostol, especially diarrhea.

   3. With the NSAID, take a PPI daily.75

   4. With the NSAID, take a full-dose histamine H2 blocker daily. The PPI and the H2 blocker reduce minor GI complaints and the risk of ulcers, but they are not rigorously proven to cut down on the serious complications, possibly because of lack of power to detect rare events in clinical trials.

Another choice that is available to reduce risk of GI events with an NSAID is to take a COX-2 selective inhibitor (“coxib”).6874 Celecoxib is the only coxib available in the United States. Because this drug does not block the “housekeeping” gene, it may not have the same GI risks as nonselective NSAIDs. The Celecoxib Long-Term Arthritis Safety Study (CLASS) study demonstrated a reduced risk of ulcer complications and symptomatic ulcers for celecoxib 400 mg daily compared with nonselective NSAID combined at the 6-month point. Celecoxib was approved and has been used extensively because of the advantages shown in this study, as well as its effectiveness, and other studies have demonstrated that it has GI advantages over nonselective NSAIDs.68

However, there remains concern that the GI protective effects of celecoxib may not be maintained long term and that it was not shown to consistently decrease the risk of serious complications (perforation, obstruction, or bleeding) as an independent endpoint. Another concern is that there is little or no gastroprotection afforded by celecoxib in those patients taking aspirin.68

Clinical Controversy…

Which is safer in OA:celecoxib alone or a nonselective NSAID with a PPI?

Celecoxib is associated with reduced risk for GI ulcers and their complications compared with nonselective NSAIDs, although this benefit is less certain with long-term use of celecoxib or for those taking aspirin. Use of PPIs with nonselective NSAIDs offers protection against development of GI ulcers, but head-to-head studies between this regimen and celecoxib are not definitive. For very high GI risk patients, the combination of celecoxib with a PPI is appealing. However, celecoxib appears to increase the risk of MI, although less so than other coxibs. Naproxen appears to be the safest NSAID to use with respect to the heart, so a difficult choice remains for the patient with high GI and cardiovascular risks.

Images Cardiovascular Risk of COX-2 Inhibitors and Traditional NSAIDs In 2004, rofecoxib was withdrawn from the market after analysis of the Adenomatous Polyp Prevention on Vioxx (APPROVe) trial, where rofecoxib doubled the risk of cardiovascular events compared with placebo.76 Celecoxib use in the Adenoma Prevention with Celecoxib (APC) trial (up to 800 mg/day) also increased cardiovascular risk.77 This prompted evaluation of all NSAIDs for possible cardiovascular risks.78 The same evaluation included meta-analyses of NSAIDs compared with placebo. Diclofenac had a significantly increased risk of cardiovascular events [RR = 1.63 (1.12 to 2.37)], while the increase seen with ibuprofen did not reach significance [RR = 1.51 (0.96 to 2.37)]. No increased risk was seen with naproxen. A drawback in this work was the small number of cardiovascular events in the meta-analyses, making it impossible to statistically differentiate the risk of one coxib compared with another.

Data from controlled trials and observational studies confirm the increased cardiovascular risk seen with rofecoxib. Celecoxib 200 mg/day or even 400 mg/day does not appear to increase risk, but cardiovascular risk is likely increased with doses above 400 mg/day. In addition, the risk of taking higher doses of celecoxib is greater for those with high cardiovascular risk than for those with low cardiovascular risk. The balance of the evidence also suggests that for the traditional NSAIDs that have been examined, with the exception of diclofenac, there is no significant or substantial increase in risk.

In a 2012 Danish study, national hospitalization records and pharmacy records of approximately 100,000 first-time MI patients revealed increased risks associated with use of any NSAID after MI. NSAID use was associated with a 1.59-fold increased risk of death (95% CI, 1.49 to 1.69) within 1 year of MI and 1.63-fold increased risk (95% CI, 1.52 to 1.74) after 5 years.79 Moreover, there was a 1.30-fold increased risk of coronary death or nonfatal recurrent MI (95% CI, 1.22 to 1.39) within 1 year of MI and 1.41-fold increased risk (95% CI, 1.28 to 1.55) after 5 years. Increased risk with NSAID use after MI was lowest with naproxen and highest with diclofenac.

Considerations for Patients at Risk for Both GI and Cardiovascular Events Images Recently, Canadian consensus guidelines were developed to recommend gastroprotection for those on NSAID therapy.73 A multidisciplinary group focused on four areas: benefits of traditional NSAIDs, aspirin, and COX-2 inhibitors; harms of traditional NSAIDs, aspirin, and COX-2 inhibitors; reducing harms of traditional NSAIDs, aspirin, and COX-2 inhibitors; and economic considerations. Recommendations were made for patients at low GI and cardiovascular risk, low GI risk and high cardiovascular risk, high GI risk and low cardiovascular risk, and for those with high GI and cardiovascular risk (Fig. 71-7).


FIGURE 71-7 Algorithm for the use of long-term NSAID therapy and gastroprotective agents according to a patient’s GI and cardiovascular risk. (Rostom A, Moayyedi P, Hunt R; Canadian Association of Gastroenterology Consensus Group. Canadian Consensus Guidelines on long-term nonsteroidal anti-inflammatory drug therapy and the need for gastroprotection: Benefits versus risks. Aliment Pharmacol Ther 2009;29(5):481. Wiley-Blackwell Publishers.)

Other Toxicities Associated with NSAIDs Images NSAIDs may cause kidney diseases, including acute renal insufficiency, tubulointerstitial nephropathy, hyperkalemia, and renal papillary necrosis.80 Clinical features of these NSAID-induced renal syndromes include increased serum creatinine and blood urea nitrogen, hyperkalemia, elevated blood pressure, peripheral edema, and weight gain. Patients at high risk are those with conditions associated with decreased renal blood flow or taking certain medications. Examples are those with chronic renal insufficiency, congestive heart failure, severe hepatic disease, and nephrotic syndrome, those of advanced age, or those taking diuretics, angiotensin-converting enzyme inhibitors, cyclosporine, or aminoglycosides (Fig. 71-8).


FIGURE 71-8 Mechanisms implicated in NSAID-induced renal injury. The minus (–) sign indicates inhibitory influence (CHF, congestive heart failure; NSAIDs, nonsteroidal antiinflammatory drugs.)

Close monitoring is advisable for high-risk patients taking an NSAID, with monitoring of serum creatinine at baseline and within 3 to 7 days of drug initiation. For those with impaired renal function, the National Kidney Foundation recommends acetaminophen over NSAIDs, although acetaminophen may pose risks, as discussed above.

Coxibs and NSAIDs uncommonly cause drug-induced hepatitis; the two NSAIDs most frequently implicated are diclofenac and sulindac. Patient monitoring should include periodic liver enzymes (aspartate aminotransferase and alanine aminotransferase), with cessation of therapy if these values exceed two to three times the upper limit of normal. In a pooled analysis of 41 studies including celecoxib, there was a low rate of serious, hepatic-related adverse events with celecoxib (1.11%), with no significant difference from naproxen or ibuprofen, but a significantly higher incidence with diclofenac (4.24%).81

Other toxic effects of NSAIDs include hypersensitivity reactions, rash, and CNS complaints of drowsiness, dizziness, headaches, depression, confusion, and tinnitus.70 It is also recommended that NSAIDs be avoided for patients with asthma who are aspirin-intolerant.

All nonspecific NSAIDs inhibit COX-1–dependent thromboxane production in platelets and thus increase bleeding risk. Unlike aspirin, celecoxib and nonspecific NSAIDs inhibit thromboxane formation reversibly, with normalization of platelet function 1 to 3 days after the drug is stopped. Warfarin and celecoxib are metabolized by the cytochrome P450 isoenzyme CYP2C9; patients receiving warfarin and COX-2 inhibitors should be followed closely.

Finally, NSAIDs should be used only with great caution and only if definitely necessary during pregnancy because of the risk to the fetus posed by the bleeding problems. In late pregnancy, all NSAIDs should be avoided.

Finally, if misoprostol is taken for GI protection, great care is indicated. Because of its abortifacient properties, misoprostol is contraindicated in pregnancy and in women of childbearing age who are not maintaining adequate contraception. It must be dispensed in its original container, which carries a warning for these individuals. Misoprostol is also available in a combination product with diclofenac, which bears the same restrictions as misoprostol alone.

Drug–Drug and Drug–Food Interactions Avoidance of concomitant use, or anticipation and careful monitoring can often prevent serious events when concomitant therapy with NSAIDs and potentially interacting drugs is being considered. The most potentially serious interactions include the use of NSAIDs with lithium, warfarin, other agents that increase bleeding risk, oral hypoglycemics, methotrexate, antihypertensives, angiotensin-converting enzyme inhibitors, β-blockers, and diuretics.70 In addition, there are probable drug interactions with tacrolimus for ibuprofen, naproxen, diclofenac, and possibly other NSAIDs.

Specific drug interactions are also seen with celecoxib.82 Celecoxib metabolism is primarily via CYP2C9.82 Cytochrome P450 inducers such as rifampin, carbamazepine, and phenytoin have the potential to reduce celecoxib levels. Concomitant administration of celecoxib with fluconazole can increase plasma concentrations of celecoxib, due to fluconazole inhibition of the CYP2C9 isoenzyme. Because warfarin and celecoxib are both metabolized by CYP2C9, patients receiving warfarin and COX-2 inhibitors should be followed closely. Celecoxib inhibits CYP2D6, and this may increase concentrations of a variety of agents, including antidepressants. Celecoxib is a sulfonamide and is thus contraindicated for those with sulfa allergies.82

Another drug interaction has been noted for those taking some NSAIDs and cardioprotective doses of aspirin. Ibuprofen at doses of 400 mg or more may block aspirin’s antiplatelet effect if it is taken first. Patients taking ibuprofen have been advised to take a single dose of ibuprofen at least 30 minutes after taking aspirin, or to take their aspirin at least 8 hours after taking ibuprofen. Other nonselective NSAIDs, such as naproxen, also may cause such interactions. Currently, the ACR recommends that patients who need an oral NSAID for OA choose an NSAID other than ibuprofen or COX-2 selective inhibitors.39

Acetaminophen does not appear to interfere with the antiplatelet effect of aspirin.

Dosing and Administration Administration of NSAIDs must be tailored to the individual patient with OA. Selection of an NSAID depends on the prescriber’s experience, medication cost, patient preference, allergies, toxicities, and adherence issues. Individual patient response differs among NSAIDs (see Table 71-2), so if an inadequate response is obtained with one NSAID, another NSAID may yet provide benefit.3739

TABLE 71-2 Drug Dosing Table



Topical NSAIDs

Pharmacology and Mechanism of Action Images The mechanism of action of topical NSAIDs is considered to be through inhibition of the COX-2 enzyme in tissues near the site of application. Studies show significant placebo effects that could result from rubbing the product into the skin, which may have a counterirritant effect. Topical NSAIDs are significantly more efficacious compared with placebo vehicle in reducing pain due to musculoskeletal conditions, including OA. Topical ketoprofen is efficacious, as well as topical ibuprofen to a certain extent, but diclofenac is the best studied and most effective topical agent. Most trials have shown topical diclofenac to be as effective as oral NSAIDs, including both oral diclofenac and other comparators.55,62,83 Diclofenac 1% gel as well as the newer diclofenac liquid drops and diclofenac patches are currently approved in the United States for OA.

Adverse Effects Compared with oral NSAIDs, topical NSAIDs are associated with many fewer GI adverse events and fewer adverse events overall, except for local application site reactions (Table 71-3).

TABLE 71-3 Drug Monitoring Table


Topical NSAIDs produce local adverse events, most often mild skin reactions such as itching or rash, but very few serious adverse effects. In a comparison of oral (n = 311) and topical diclofenac (n = 311), significantly more persons receiving topical diclofenac developed dry skin, rash, and itching, though none was considered serious. However, significantly more persons receiving oral diclofenac had severe GI effects, asthma, dizziness, dyspnea, change from normal to abnormal hemoglobin, ALT increase to more than three times upper limit of normal, and creatinine clearance changing from normal to abnormal.55

Case–control studies revealed that hospital admissions for upper GI bleeding and perforation in those who had used topical NSAIDs in the prior 45 days, when adjusted for oral NSAID use and ulcer healing drugs, was not significantly increased relative to either community or hospital controls. The relative risk for those who had received oral NSAIDs was significantly increased by 2.6-fold and 2.0-fold using community or hospital controls, respectively.55 A nested case–control study from the United Kingdom revealed no significant association between topical NSAID use and renal failure, whereas oral NSAID use was significantly associated with a doubling of risk.55

An estimated 1% to 15% of topical NSAID enters the systemic circulation (usually less than 5%), and this contributes to its favorable safety profile.38,83

Drug–Drug Interactions Interactions listed for topical diclofenac are the same as those listed above for oral NSAIDs. The most potentially serious interactions include the use of NSAIDs with lithium, warfarin, and other agents that increase bleeding risk, oral hypoglycemics, methotrexate, antihypertensives, angiotensin-converting enzyme inhibitors, β-blockers, and diuretics. Other topical agents have not been studied with topical diclofenac, and changes in tolerability and absorption are possible.

For all of these interactions, as there is only a small percentage of diclofenac absorbed, the risks are likely significantly less than with oral drug, but the patient and provider would be wise to monitor appropriately for these interactions for any of these drugs the patient is taking.

Patients should avoid oral NSAIDs while using topical products to minimize potential for additive adverse effects. Care should be taken to avoid contact with the eyes or open wounds and to wash hands after application (except when treating hand OA).

Dosing and Administration Diclofenac 1% gel (Voltaren) can be used for hand or knee OA or other joints amenable to topical application (not the hip). The gel is applied four times daily using the dose measuring cards provided by the manufacturer. For application to the affected area in the lower limbs, the recommended dose is 4 g four times daily. For upper extremities, the recommended dose is 2 g four times daily.

Diclofenac drops (Pennsaid), only approved for knee OA, are provided in a 1.5% solution in dimethylsulfoxide, and 40 drops are applied four times a day for each knee to be treated. The solution should be applied to the back, front, and sides of the knee. For each dose, the patient places 10 drops at a time directly onto the painful knee (or first into the hands and then immediately onto the knee) and rubs the solution in. The patient then repeats this process three more times until 40 drops have been applied to the painful knee for that particular dose.

The diclofenac patch (diclofenac epolamine 180 mg) is applied twice daily. If the patch does not stick well, the patient can tape edges with first-aid bandages. Patient counseling is important to carefully explain how to apply the topical products and how long to wait before dressing, putting on gloves, showering, and so forth.

Intraarticular Corticosteroids

Images Pharmacology and Mechanism of Action The antiinflammatory properties of corticosteroids as a class are the primary mechanism of pain relief in the treatment of OA. These properties decrease the formation and release of prostaglandins, kinins, liposomal enzymes, and histamine. Intraarticular corticosteroid injections can provide excellent pain relief, particularly when a joint effusion is present.39,84

Aspiration of the effusion and injection of glucocorticoid are carried out aseptically, with examination of the aspirate recommended to exclude crystalline arthritis or infection. The incidence of infection is low, however—approximately 1 in 50,000 procedures.

Several randomized, placebo-controlled, double-blind studies have shown that intraarticular corticosteroids are superior to placebo in alleviating knee pain and stiffness caused by OA.84 The branched esters of triamcinolone and methylprednisolone are preferred by practitioners because of the reduced solubility that allows the agents to remain in the joint space longer. There is no evidence of a clinically superior corticosteroid for intraarticular use, with equipotent doses of methylprednisolone acetate and triamcinolone hexacetonide having similar efficacy.85

Adverse Events Adverse events associated with intraarticular injection of corticosteroids can be local or systemic in nature.

Systemic adverse events are the same as with any other systemic corticosteroid and can include hyperglycemia, edema, elevated blood pressure, flushing, dyspepsia, and adrenal suppression. Multiple injections over longer periods of time (up to 10 years) are more likely to lead to adrenal suppression as recovery time to baseline adrenal function is longer with repeated injections.86 Hyperglycemia may occur in patients with stable diabetes mellitus as well as those without history of abnormal glycemic control.86 Hyperglycemia can occur within 24 hours of injection and last for up to 2 weeks.

Local adverse effects can include infection in the affected joint, osteonecrosis, tendon rupture, and skin atrophy at the injection site. It has long been thought that intraarticular corticosteroids can hasten cartilage loss, but the potential risk of cartilage destruction with steroid injections has not been substantiated. Systemic corticosteroid therapy is not recommended in OA, given the lack of proven benefit and the well-known adverse effects with long-term use.

Dosing and Administration Average doses for injection of large joints in adults are 10 to 20 mg of triamcinolone hexacetonide or 20 to 40 mg of methylprednisolone acetate. This therapy is generally limited to three or four injections per year due to the potential systemic effects of corticosteroids and because the need for more frequent injections indicates little response to the therapy.

After injection, the patient should minimize activity and stress on the joint for several days. Initial pain relief may be seen within 24 to 72 hours after injection, with peak pain relief about 7 to 10 days after injection and lasting up to 4 to 8 weeks.


Pharmacology and Mechanism of Action Capsaicin, isolated from hot peppers, releases and ultimately depletes substance P from afferent nociceptive nerve fibers. Substance P has been implicated in the transmission of pain in arthritis, and capsaicin cream has been shown in four placebo-controlled studies to provide pain relief in knee and hand OA when applied over affected joints.62 Due to the larger surface area and distance from the site of application to the joint, it is not expected that application of capsaicin would provide efficacy in the treatment of hip OA.

Adverse Effects Adverse events associated with capsaicin are primarily local, with one in three patients experiencing burning, stinging, and/or erythema that usually subsides with repeated application. FDA has recently issued a public drug safety communication notifying consumers that rare cases of severe burns have been reported.87 Some patients may experience coughing associated with application.

Dosing and Administration To be effective, capsaicin must be used regularly, and it may take up to 2 weeks to take effect. Although use is recommended four times a day, a twice-daily application may enhance long-term adherence and still provide adequate pain relief.62 Patients should be counseled not to get the cream in their eyes or mouth. Patients should also notify their healthcare provider immediately if they experience pain, swelling, or blistering skin at the site of application.

When patients were queried using an electronic questionnaire that considered possible toxicities of treatments, as well as route of administration and cost, capsaicin was the most preferred by patients, even when it was portrayed as being less effective than NSAIDs.88

Capsaicin is a nonprescription product available as a cream, gel, or lotion in concentrations ranging from 0.025% to 0.075%.


Pharmacology and Mechanism of Action Tramadol, an analgesic with affinity for the μ-opioid receptor, as well as weak inhibition of the reuptake of norepinephrine and serotonin neurotransmitter, has modest analgesic effects (with or without acetaminophen) for patients with OA when compared with placebo.89 Tramadol is also modestly effective as add-on therapy for patients taking concomitant acetaminophen, NSAIDs or COX-2–selective inhibitors. Tramadol may be helpful for patients who cannot take NSAIDs or COX-2–selective inhibitors.

Adverse events Opioid-like adverse effects such as nausea, vomiting, dizziness, constipation, headache, and somnolence are common with tramadol. These occur in 60% to 70% of treated patients, and 40% discontinue tramadol because of an adverse effect.89 Although the frequency of adverse effects is high, the severity of adverse effects is less than with NSAIDs, as tramadol use is not associated with life-threatening GI bleeding, cardiovascular events, or renal failure. The most notable serious adverse event associated with tramadol use is seizures. Tramadol is not classified as a controlled substance, but there are numerous reports of patients displaying drug-seeking behaviors similar to those displayed by patients who are dependent and/or addicted to opioid analgesics.39

Drug–Drug Interaction Medications that lower the seizure threshold should be used with caution in patients taking tramadol. These include tricyclic antidepressants, first-generation antipsychotic medications and cyclobenzaprine, as well as others. There is also an increased risk of serotonin syndrome (see eChap. 10, Clinical Toxicology, for description and management of this condition) when tramadol is used concomitantly with other serotonergic medications, including duloxetine.

Dosing and Administration Tramadol should be initiated at a lower dose (100 mg per day) and may be titrated as needed for pain control to a dose of 200 mg per day.

Tramadol is available in a combination tablet with acetaminophen and as a sustained-release (SR) tablet.

Second-Line Treatments

Opioid Analgesics

Opioid analgesics can be useful for patients who experience limited pain relief with acetaminophen, oral NSAIDs, intraarticular injections, or topical therapy.90 For patients with underlying conditions that limit the use of first-line analgesics, opioid analgesics can effectively relieve acute OA pain. A common clinical scenario may include the patient who cannot take oral NSAIDs because of renal failure or cardiovascular disease. Patients in whom all other treatment options have failed and who are at high surgical risk, precluding joint arthroplasty are also candidates for opioid therapy. As many patients with OA are elderly, it is important to carefully use opioids to promote safety. The following recommendations have been suggested to optimize opioid therapy: (a) use the least invasive route of administration, (b) initiate one agent at a time, at a low dose, (c) allow a sufficiently long interval between dose increases to allow an assessment of efficacy and safety, (d) use a long-acting preparation, (e) therapy should be constantly monitored and adjusted if necessary, (f) changing opioids may be necessary.90

SR compounds usually offer better pain control throughout the day, and are used when immediate-release (IR) opioids do not provide a sufficient duration of pain control. A variety of IR and SR opioid compounds have been studied including oxycodone IR and SR, morphine IR and SR, hydromorphone, and fentanyl transdermal patch.57

Adverse effects are common in opioid-treated OA patients. More than 75% of patients in clinical trials experience at least one typical opioid-related (i.e. nausea, somnolence, constipation, dry mouth, and dizziness) adverse effect. Although this is not an unexpected finding, it serves as a reminder to use opioids cautiously in elderly patients who may be more susceptible to adverse effects.

Opioid dependence, addiction, tolerance, hyperalgesia, and issues surrounding drug diversion are more serious adverse effects associated with long-term treatment. Prescription opioid misuse/abuse/addiction is a major public health concern with the CDC reporting almost 15,000 deaths in 2008.91 In 2009, there were 475,000 emergency room visits attributed to the misuse and abuse of prescription opioids.91 To address this growing safety issue, the American Pain Society/American Academy of Pain Medicine has published recommendations on the use of opioids in the management of chronic noncancer pain.92

If pain is intolerable and limits activities of daily living, and the patient has sufficiently good cardiopulmonary health to undergo major surgery, joint replacement may be preferable to continued reliance on opioids.


Duloxetine is a centrally acting dual-reuptake inhibitor of both serotonin and norepinephrine, although norepinephrine reuptake inhibition does not occur until doses reach 60 mg/day. While the most common pain target in OA is peripheral nociceptive pain, there is some evidence that chronic nociceptive pain leads to central pain sensitization, thereby lowering the pain threshold.59 Duloxetine provides pain relief through blockade of central pain transmitters, including serotonin and norepinephrine.

Adverse effects commonly associated with duloxetine therapy include nausea, dry mouth, constipation, and anorexia. Expected neurologic adverse effects include fatigue, somnolence, and dizziness. Rare but serious adverse events associated with duloxetine include Stevens–Johnson syndrome and liver failure. Patients should contact their healthcare provider immediately if they develop a rash while taking duloxetine.

Particular care should be taken to avoid the use of duloxetine with other serotonergic medications, including tramadol. As tramadol is a first-line treatment recommendation for OA, the likelihood of encountering this combination is high. Concomitant use of duloxetine with other medications that increase serotonin concentrations increases the risk of serotonin syndrome (see eChap. 10, Clinical Toxicology).

Hyaluronic Acid Injections

Agents containing HA (sodium hyaluronate) are available for intraarticular injection for treatment of knee OA.93

High-molecular-weight HA is an important constituent of synovial fluid. Endogenous HA may also have antiinflammatory effects. Because the concentration and molecular size of synovial HA decreases in OA, administration of exogenous HA products have been studied, with the theory that this could reconstitute synovial fluid and reduce symptoms. In fact, HA injections temporarily and modestly increase viscosity. When evaluating pain score improvements, it is essential to determine if score improvement corresponds to clinically meaningful improvement for patients. Extensive evaluation of the literature has revealed potential publication bias of HA studies including bias related to high levels of industry sponsorship as well as a substantial number of studies with unpublished data.94

Most HA products are injected once weekly for either 3 or 5 weeks, depending on the specific agent administered. Injections are generally well tolerated, although acute joint swelling, effusion, and stiffness can occur as well as local skin reactions, including rash, ecchymoses, and pruritus have been reported. Rarely, systemic adverse events including hypersensitivity reactions have occurred.

HA injections have limited beneficial effects for patients with knee OA. HA products have not been shown to benefit patients with hip OA.95 These agents are expensive because the treatment includes both drug costs and administration costs.

Glucosamine and Chondroitin

Interest in chondroitin and glucosamine was spurred initially by anecdotal reports of benefit in animals and humans and by the ability of these substances to stimulate proteoglycan synthesis from articular cartilage in vitro. Over the last decade, enthusiasm for these agents has waned as additional efficacy data have become available to the point that the American College of Rheumatology conditionally recommends against the use of glucosamine and chondroitin.39 Glucosamine, alone or in combination, has not been shown to provide uniform improvements in pain control or functional status in patients with OA of the knee or hip.96

Numerous trials have been conducted with suboptimal study designs, and this has led to a variable response to glucosamine and/or chondroitin.94 In contrast to the early reports of suboptimally designed studies, a large, well-controlled National Institutes of Health-sponsored study demonstrated no significant clinical response to glucosamine therapy alone, chondroitin therapy alone, or combination glucosamine–chondroitin therapy when compared with placebo across all patients.97 This trial provided high-quality evidence on the use of glucosamine and chondroitin in OA and demonstrated that the safety and efficacy glucosamine and chondroitin therapy was similar to placebo.

Because glucosamine and chondroitin are marketed in the United States as dietary supplements, neither the products nor their purity is adequately regulated by the FDA. The potential consequences related to the lack of regulatory oversight for these products can affect both efficacy and safety. Products containing less than labeled doses can compromise efficacy, while those containing ingredients not included on the labeling can compromise safety. A variety of brand name and generic products are available.

Considerations for Future Therapeutic Options

Strategies aimed at expanding therapeutic options for OA include an array of disease-modifying drugs, new drug classes to provide symptomatic relief of OA pain, and behavior modification strategies to improve patient participation in nonpharmacologic therapies.98

Disease-modifying drugs are targeted at preventing, retarding, or reversing damage to articular cartilage. Currently, OA is a progressive disease. Current approaches to slow progression of OA are directed at three different tissue-specific targets: (a) cartilage, (b) synovial membrane and associated inflammation, and (c) subchondral bone. Therapies directed at preserving cartilage include enzyme inhibitors of MMPs, inhibitors of inducible nitric oxide synthase, cathepsin K inhibitors, and nerve growth factor inhibitors.98 Several of these investigational agents are in Phase I and Phase II clinical trials in humans. Doxycycline, as a TIMP, potentially decreases cartilage destruction. In knee OA, doxycycline has been shown to delay loss of articular cartilage (joint space narrowing) in humans when compared with placebo, although the clinical impact of this finding is unclear.99

Several antiinflammatory agents targeting symptom improvement as well as structure-modifying properties at the synovial membrane are in clinical trials. The agents include interleukin-1 inhibitors, the antitumor necrosis factor inhibitor adalimumab, and adenosine A2 and A3 receptor agonists.98 Current animal research supports these receptor targets to prevent ongoing joint destruction, and early results for some of these agents, particularly adalimumab, are encouraging.

Ongoing phase III clinical trials with licofelone, a lipoxygenase/cyclooxygenase inhibitor (LOX/COX), has demonstrated some preliminary data in treatment of OA pain. The efficacy of licofelone may be similar to that of NSAIDs, but adverse event advantages are still being investigated.100 Combined LOX/COX inhibition may decrease the production of proinflammatory leukotrienes associated with the progression of OA.

Slowing the progression of OA may also be achieved by attempts to modify or repair bony changes associated with OA. Current strategies being evaluated in humans include the use of bisphosphonates, calcitonin, cholecalciferol, selective estrogen receptor modulators, parathyroid hormone, strontium, and MMP-13 inhibitors. The definitive role of these agents in modifying bone resorption as a strategy to delay the progression of bone damage associated with OA is yet to be determined.

Additionally, new agents and methods to treat symptoms of OA are being studied. These approaches include nerve growth factor inhibitors, cannabinoid receptor agonists, bradykinin receptor antagonists, kainate receptor antagonists, and transient receptor potential ion channel agonists (TRVP-1).98

Of these agents, the nerve growth inhibitor tanezumab has undergone the most extensive evaluation. Unfortunately, Phase III study results with tanezumab have revealed potential safety issues with treatment, and it appears unlikely to become a viable treatment for OA pain.101

Several other compounds targeted toward the described receptors are under active investigation. Although many of these compounds are years away from potential market approval, the extensive nature of the work is encouraging.

In addition to pharmacologic agents, acupuncture has been examined in OA. In a systematic analysis of 18 randomized, controlled trials of manual or electroacupuncture, 10 showed positive effects for acupuncture.102 However, in a recent, large, randomized, and well-controlled study, acupuncture was not seen to be any more effective than sham controls.103


OA has substantial negative impact on the quality of life for individual patients. OA is also associated with a negative impact on society as the disease is extremely common, and OA ranks second in causes of disability in the United States.39

Most OA patients use a multidisciplinary approach to their treatment.39,104 Treatments include nonpharmacologic and pharmacologic therapy, in addition to surgical options in some patients. Unfortunately, many patients have less than optimal response to treatment and commonly require a change in therapy or augmentation of partially effective therapy. Achieving adequate pain control and minimizing functional impairment in OA patients require careful assessment of comorbid conditions in each patient to safely provide effective pharmacotherapy treatments. Nonpharmacologic interventions may also require regular reinforcement and modifications.

A multidisciplinary intervention for knee OA initiated by pharmacists has been shown to improve adherence to OA guideline recommendations, decrease pain scores, and improve functional assessment scores.105 These types of multidisciplinary disease management programs that implement strategies to provide comprehensive care should be offered to all OA patients to maximize outcomes.

Total indirect and direct medical costs for OA patients are high.104 The highest costs associated with the pharmacotherapy of OA are hospitalization for treatment of NSAID-related complications, particularly serious GI adverse events. Historically, gastroprotective therapy or the use of COX-2–selective inhibitors for low-risk patients has not been cost-effective because of the large number needed to treat to prevent serious events, but current data suggest that concomitant PPI therapy in low-risk patients is cost-effective if the agent selected is a generic, multisource product.106 The use of COX-2–selective inhibitors to protect gastric mucosa in aspirin users is not cost-effective, because aspirin negates most, if not all, of the gastroprotective effects of these agents.107 Pharmacoeconomic considerations for OA involve the proper selection of therapy for the initial treatment of each patient with OA. Use of the nonprescription analgesic acetaminophen as initial therapy has greatly reduced medication costs in comparison with the use of NSAIDs, many of which are by prescription only. Oral NSAID costs vary considerably, depending on the medication, daily dose, and regimen selected. As oral NSAIDs as a class are therapeutically similar, the use of a less-expensive agent such as nonprescription ibuprofen or naproxen or a multisource generic product may minimize the cost. More-expensive NSAIDs can be prescribed if neither of these offers benefit after a 2-week trial at sufficient doses. Topical NSAIDs are significantly more costly than oral agents, although may still be cost-effective in patients at high-risk for costly complications associated with oral NSAID therapy.


For the person with OA, treatment decisions and pharmacotherapy monitoring is patient specific. The patient’s situation and individual needs should be considered when devising a treatment plan. Is the patient bothered primarily by pain, by limitations in activity, or with concerns about side effects from medications? Does the patient understand what OA is and why certain treatments are useful?

When the patient is first being assessed for the possibility of OA, the diagnosis is often straightforward, including history and physical exam, plain films of the affected joint(s), and laboratory tests. The older patient with unilateral knee pain, limited range of motion, no palpable warmth, crepitus, without prolonged morning stiffness, and without other suspicious findings is highly likely to have knee OA. It is still reasonable to obtain x-ray films, which may help follow disease over time (although joint space narrowing often does not correlate with the extent of pain or difficulty walking). Basic laboratory tests can shed light on what pharmacologic therapy is possible (e.g., in a patient with poor renal function, NSAIDs should be avoided). Pain can be assessed using a visual analog scale, and physical examination is helpful to determine range of motion for affected joints. Additional tests of OA severity may include measurement of grip strength, 50 ft walking time, patient and physician global assessment of OA severity, and assessment of ability to perform activities of daily living. Once the patient is assessed and diagnosed, patient and family education is essential. Nondrug therapy may include a referral for physical and/or occupational therapy services, where the therapists can help to maintain and improve range of motion, decrease pain modestly, lose weight if necessary, and become more active once the OA symptoms are better controlled.

Setting the stage for pharmacotherapy with the above is important but in the meantime, the patient needs pain relief. A few years ago, acetaminophen would be the only “first choice” for pain relief. Adverse events with acetaminophen are uncommon, although it is crucial that the patient understand the maximum daily dose limits and realize that many products contain acetaminophen. Although some patients do well on acetaminophen, many do not achieve sufficient pain relief.

A step up to oral NSAIDs or opioid therapy might be necessary, but this poses significant risks beyond acetaminophen. A switch to NSAIDs requires careful consideration of the patient’s age and comorbidities, renal function, history of GI problems, hypertension, and cardiovascular health. Periodic monitoring would include open-ended questions followed by direct questions relating to the commonest adverse effects associated with the respective medication. For an oral NSAID, symptoms of abdominal pain, heartburn, nausea, or change in stool color provide valuable clues to the presence of GI complications, although serious GI complications can occur without warning. Patients should be monitored for the development of hypertension, weight gain, edema, skin rash, and CNS adverse effects such as headaches and drowsiness. Baseline serum creatinine, complete blood count, and serum transaminases are repeated at 6- to 12-month intervals to identify GI, renal, and hepatic toxicities.

Topical NSAIDs are now known to have efficacy in OA of the hand and knee and are as effective as oral NSAIDs. Although they carry the same cardiovascular, renal, and GI warnings, the amount of a typical dose absorbed into the bloodstream is only a few percent of that from an equivalent dose of oral NSAID (as measured by areas under serum concentration–time curves). Topical NSAIDs’ most common adverse effects are local, with irritated skin, rash, or itching, usually mild, and with many fewer adverse effects of cardiovascular, GI, or renal nature. These agents are a welcome addition to the limited treatment modalities for the very common, costly, painful, and often disabling disease of OA. It is important that the patient apply the topical products appropriately to achieve maximum benefit and avoiding adverse events.

For patients receiving intraarticular corticosteroids, improvement should begin within 2 to 3 days and last 4 to 8 weeks. Patients should be advised about possible injection site reactions, as well as possible systemic effects, especially for those with hypertension or diabetes, as there is a potential for increased blood pressure or blood glucose. For patients receiving opioids or tramadol, relief from pain should occur rapidly. Frail or elderly patients should be monitored carefully and cautioned about sedation, dysphoria, nausea, risk of falls, and constipation. Additional monitoring should include strategies to assess development of opioid tolerance and addiction.


OA is a very common, slowly progressive disorder that affects diarthrodial joints and is characterized by progressive deterioration of articular cartilage, subchondral sclerosis, and osteophyte production. Clinical manifestations include gradual onset of joint pain, stiffness, and limitation of motion. The primary treatment goals are to reduce pain, maintain function, and prevent further destruction. An individualized approach based on education, rest, exercise, weight loss as needed, and analgesic medication can succeed in meeting these goals. Recommended drug treatment starts with acetaminophen ≤4 g/day and topical analgesics as needed. If acetaminophen is ineffective, oral NSAIDs may be used in appropriately selected patients, often providing satisfactory relief of pain and stiffness. Individuals at increased risk for toxicity from NSAIDs, especially for GI, cardiovascular, or renal events, deserve special attention. Celecoxib may have safety advantages in some OA patients, but its safety relative to other NSAIDs and its role in OA remains poorly defined. Adjunctive therapy with tramadol, intraarticular corticosteroids, and opioid analgesics may be helpful in patients with poorly controlled pain. Experimental therapy aimed at preventing the progression of OA requires further clinical investigation before entering widespread clinical use.




    1. Zhang Y, Jordan JM. Epidemiology of osteoarthritis. Clin Geriatr Med 2010;26(3):355–369.

    2. Bijlsma JW, Berenbaum F, Lafeber FP. Osteoarthritis: An update with relevance for clinical practice. Lancet 2011;377(9783):2115–2126.

    3. Murphy L, Helmick CG. The impact of osteoarthritis in the United States: A population-health perspective. Am J Nurs 2012;112(3 Suppl 1):S13–S19.

    4. Lawrence RC, Felson DT, Helmick CG, et al. Estimates of the prevalence of arthritis and other rheumatic conditions in the United States. Part II. Arthritis Rheum 2008;58(1):26–35.

    5. Hunter DJ. Insights from imaging on the epidemiology and pathophysiology of osteoarthritis. Radiol Clin North Am 2009;47(4):539–551.

    6. Chu CR, Williams AA, Coyle CH, Bowers ME. Early diagnosis to enable early treatment of pre-osteoarthritis. Arthritis Res Ther 2012;14(3):212.

    7. Kotlarz H, Gunnarsson CL, Fang H, Rizzo JA. Osteoarthritis and absenteeism costs: Evidence from US national survey data. J Occup Environ Med 2010;52(3):263–268.

    8. Dagenais S, Garbedian S, Wai EK. Systematic review of the prevalence of radiographic primary hip osteoarthritis. Clin Orthop Relat Res 2009;467(3):623–637.

    9. Feydy A, Pluot E, Guerini H, Drape JL. Osteoarthritis of the wrist and hand, and spine. Radiol Clin North Am 2009;47(4):723–759.

   10. Hunter DJ. In the clinic: Osteoarthritis. Ann Intern Med 2007;147(3):ITC8-1–ITC8-16.

   11. Kopec JA, Rahman MM, Sayre EC, et al. Trends in physician-diagnosed osteoarthritis incidence in an administrative database in British Columbia, Canada, 1996–1997 through 2003–2004. Arthritis Rheum 2008;59(7):929–934.

   12. Garstang SV, Stitik TP. Osteoarthritis: Epidemiology, risk factors, and pathophysiology. Am J Phys Med Rehabil 2006;85(11 Suppl):S2–11; quiz S12–S14.

   13. Mork PJ, Holtermann A, Nilsen TI. Effect of body mass index and physical exercise on risk of knee and hip osteoarthritis: Longitudinal data from the Norwegian HUNT study. J Epidemiol Community Health 2012;66(8):678–683.

   14. Andersen S, Thygesen LC, Davidsen M, Helweg-Larsen K. Cumulative years in occupation and the risk of hip or knee osteoarthritis in men and women: A register-based follow-up study. Occup Environ Med 2012;69(5):325–330.

   15. Sun BH, Wu CW, Kalunian KC. New developments in osteoarthritis. Rheum Dis Clin North Am 2007;33(1):135–148.

   16. Botha-Scheepers SA, Watt I, Slagboom E, et al. Influence of familial factors on radiologic disease progression over two years in siblings with osteoarthritis at multiple sites: A prospective longitudinal cohort study. Arthritis Rheum 2007;57(4):626–632.

   17. Valdes AM, Spector TD. The contribution of genes to osteoarthritis. Med Clin North Am 2009;93(1):45–66, x.

   18. Valdes AM, Spector TD. Genetic epidemiology of hip and knee osteoarthritis. Nat Rev Rheumatol 2011;7(1):23–32.

   19. Evangelou E, Valdes AM, Kerkhof HJ, et al. Meta-analysis of genome-wide association studies confirms a susceptibility locus for knee osteoarthritis on chromosome 7q22. Ann Rheum Dis 2011;70(2):349–355.

   20. Meulenbelt I. Osteoarthritis year 2011 in review: Genetics. Osteoarthritis Cartilage 2012;20(3):218–222.

   21. Dieppe P. Developments in osteoarthritis. Rheumatology (Oxford) 2011;50(2):245–247.

   22. Sandell LJ, Heinegard D, Hering TH. Cell biology, biochemistry, and molecular biology of articular cartilage in osteoarthritis. In: Moscowitz RW, Altman RD, Hochberg MC, Buckwalter JA, Goldberg VM, eds. Osteoarthritis: Diagnosis and Medical/Surgical Management, 4th ed. Philadelphia: Lippincott, Williams, & Wilkins, 2007:73–106.

   23. Felson DT. Osteoarthritis. In: Fauci AS, Braunwald E, Kaspar DL, Hauser DL, Longo DLJ, Loscalzo J, eds. Harrison’s Principles of Internal Medicine. New York, NY: McGraw-Hill, 2009:2158–2165.

   24. Loeser RF, Goldring SR, Scanzello CR, Goldring MB. Osteoarthritis: A disease of the joint as an organ. Arthritis Rheum 2012;64(6):1697–1707.

   25. Goldring SR, Goldring MB. The role of cytokines in cartilage matrix degeneration in osteoarthritis. Clin Orthop Relat Res 2004;(427 Suppl):S27–S36.

   26. Vincent TL, Saklatvala J. Is the response of cartilage to injury relevant to osteoarthritis? Arthritis Rheum 2008;58(5):1207–1210.

   27. Goldring MB, Marcu KB. Cartilage homeostasis in health and rheumatic diseases. Arthritis Res Ther 2009;11(3):224.

   28. Dell’accio F, De Bari C, Eltawil NM, Vanhummelen P, Pitzalis C. Identification of the molecular response of articular cartilage to injury, by microarray screening: Wnt-16 expression and signaling after injury and in osteoarthritis. Arthritis Rheum 2008;58(5):1410–1421.

   29. Fukui N, Ikeda Y, Ohnuki T, et al. Regional differences in chondrocyte metabolism in osteoarthritis: A detailed analysis by laser capture microdissection. Arthritis Rheum 2008;58(1):154–163.

   30. Wu J, Liu W, Bemis A, et al. Comparative proteomic characterization of articular cartilage tissue from normal donors and patients with osteoarthritis. Arthritis Rheum 2007;56(11):3675–3684.

   31. Karsdal MA, Leeming DJ, Dam EB, et al. Should subchondral bone turnover be targeted when treating osteoarthritis? Osteoarthritis Cartilage 2008;16(6):638–646.

   32. Rosenberg AE. Bones, joints and soft tissue tumors. In: Kumar V, Abbas AK, Fausto N, Aster J, eds. Robbins and Cotran Pathologic Basis of Disease, Professional Edition, 8th ed. Philadelphia, PA: W.B. Saunders, 2009:1235–1236.

   33. Altman RD. Laboratory diagnosis of osteoarthritis. In: Moscowitz RW, Altman RD, Hochberg MC, Buckwalter JA, Goldberg VM, eds. Osteoarthritis: Diagnosis and Medical/Surgical Management, 4th ed. Philadelphia: Lippincott, Williams, & Wilkins, 2007:201–214.

   34. Lane NE. Clinical practice. Osteoarthritis of the hip. N Engl J Med 2007;357(14):1413–1421.


   36. Zhang W, Doherty M, Leeb BF, et al. EULAR evidence based recommendations for the management of hand osteoarthritis: Report of a task force of the EULAR standing committee for international clinical studies including therapeutics (ESCISIT). Ann Rheum Dis 2007;66(3):377–388.

   37. Bennell KL, Hunter DJ, Hinman RS. Management of osteoarthritis of the knee. BMJ 2012;345:e4934.

   38. Conaghan PG, Dickson J, Grant RL, Guideline Development Group. Care and management of osteoarthritis in adults: Summary of NICE guidance. BMJ 2008;336(7642):502–503.

   39. Hochberg MC, Altman RD, April KT, et al. American college of rheumatology 2012 recommendations for the use of nonpharmacologic and pharmacologic therapies in osteoarthritis of the hand, hip, and knee. Arthritis Care Res (Hoboken) 2012;64(4):455–474.

   40. Knittle K, De Gucht V, Maes S. Lifestyle- and behaviour-change interventions in musculoskeletal conditions. Best Pract Res Clin Rheumatol 2012;26(3):293–304.

   41. Hawker GA, Mian S, Bednis K, Stanaitis I. Osteoarthritis year 2010 in review: Non-pharmacologic therapy. Osteoarthritis Cartilage 2011;19(4):366–374.

   42. Christensen R, Bartels EM, Astrup A, Bliddal H. Effect of weight reduction in obese patients diagnosed with knee osteoarthritis: A systematic review and meta-analysis. Ann Rheum Dis 2007;66(4):433–439.

   43. Reijman M, Pols HA, Bergink AP, et al. Body mass index associated with onset and progression of osteoarthritis of the knee but not of the hip: The Rotterdam study. Ann Rheum Dis 2007;66(2):158–162.

   44. Vincent HK, DeJong G, Mascarenas D, Vincent KR. The effect of body mass index and hip abductor brace use on inpatient rehabilitation outcomes after total hip arthroplasty. Am J Phys Med Rehabil 2009;88(3):201–209.

   45. Messier SP. Diet and exercise for obese adults with knee osteoarthritis. Clin Geriatr Med 2010;26(3):461–477.

   46. Jenkinson CM, Doherty M, Avery AJ, et al. Effects of dietary intervention and quadriceps strengthening exercises on pain and function in overweight people with knee pain: Randomised controlled trial. BMJ 2009;339:b3170.

   47. Josbeno DA, Kalarchian M, Sparto PJ, Otto AD, Jakicic JM. Physical activity and physical function in individuals post-bariatric surgery. Obes Surg 2011;21(8):1243–1249.

   48. Messier SP, Legault C, Mihalko S, et al. The intensive diet and exercise for arthritis (IDEA) trial: Design and rationale. BMC Musculoskelet Disord 2009;10:93.

   49. Fernandes L, Storheim K, Nordsletten L, Risberg MA. Development of a therapeutic exercise program for patients with osteoarthritis of the hip. Phys Ther 2010;90(4):592–601.

   50. Fitzpatrick R, Shortall E, Sculpher M, et al. Primary total hip replacement surgery: A systematic review of outcomes and modelling of cost-effectiveness associated with different prostheses. Health Technol Assess 1998;2(20):1–64.

   51. Gossec L, Paternotte S, Maillefert JF, et al. The role of pain and functional impairment in the decision to recommend total joint replacement in hip and knee osteoarthritis: An international cross-sectional study of 1909 patients. Report of the OARSI-OMERACT task force on total joint replacement. Osteoarthritis Cartilage 2011;19(2):147–154.

   52. Skou ST, Roos EM, Laursen MB, et al. Total knee replacement plus physical and medical therapy or treatment with physical and medical therapy alone: A randomised controlled trial in patients with knee osteoarthritis (the MEDIC-study). BMC Musculoskelet Disord 2012;13:67.

   53. Losina E, Walensky RP, Kessler CL, et al. Cost-effectiveness of total knee arthroplasty in the United States: Patient risk and hospital volume. Arch Intern Med 2009;169(12):1113–1121; discussion 1121–1122.

   54. Kuo CK, Li WJ, Mauck RL, Tuan RS. Cartilage tissue engineering: Its potential and uses. Curr Opin Rheumatol 2006;18(1):64–73.

   55. Chou R, McDonagh MS, Nakamoto E, Griffin J. Analgesics for Osteoarthritis: An Update of the 2006 Comparative Effectiveness Review. Comparative Effectiveness Review No. 38 (Prepared by the Oregon Evidence-based Practice Center under Contract HHSA 290 2007 10057 I) AHRQ Publication No. 11(12)-EHC076-EF. Rockville, MD: Agency for Healthcare Research and Quality, October 2011.

   56. Towheed TE, Maxwell L, Judd MG, Catton M, Hochberg MC, Wells G. Acetaminophen for osteoarthritis. Cochrane Database Syst Rev 2006;(1):CD004257.

   57. Nuesch E, Rutjes AW, Husni E, Welch V, Juni P. Oral or transdermal opioids for osteoarthritis of the knee or hip. Cochrane Database Syst Rev 2009;(4):CD003115.

   58. Frakes EP, Risser RC, Ball TD, Hochberg MC, Wohlreich MM. Duloxetine added to oral nonsteroidal anti-inflammatory drugs for treatment of knee pain due to osteoarthritis: Results of a randomized, double-blind, placebo-controlled trial. Curr Med Res Opin 2011;27(12):2361–2372.

   59. Hochberg MC, Wohlreich M, Gaynor P, Hanna S, Risser R. Clinically relevant outcomes based on analysis of pooled data from 2 trials of duloxetine in patients with knee osteoarthritis. J Rheumatol 2012;39(2):352–358.

   60. Altman RD, Dreiser RL, Fisher CL, Chase WF, Dreher DS, Zacher J. Diclofenac sodium gel in patients with primary hand osteoarthritis: A randomized, double-blind, placebo-controlled trial. J Rheumatol 2009;36(9):1991–1999.

   61. Altman RD. Pharmacological therapies for osteoarthritis of the hand: A review of the evidence. Drugs Aging 2010;27(9):729–745.

   62. Altman RD, Barthel HR. Topical therapies for osteoarthritis. Drugs 2011;71(10):1259–1279.

   63. Henrich WL, Agodoa LE, Barrett B, et al. Analgesics and the kidney: Summary and recommendations to the scientific advisory board of the national kidney foundation from an ad hoc committee of the National Kidney Foundation. Am J Kidney Dis 1996;27(1):162–165.

   64. Kuo HW, Tsai SS, Tiao MM, Liu YC, Lee IM, Yang CY. Analgesic use and the risk for progression of chronic kidney disease. Pharmacoepidemiol Drug Saf 2010;19(7):745–751.

   65. Federal Register. Rockville, MD: Department of Health and Human Services Vol. 74, No. 81, 2009.

   66. Watkins PB, Kaplowitz N, Slattery JT, et al. Aminotransferase elevations in healthy adults receiving 4 grams of acetaminophen daily: A randomized controlled trial. JAMA 2006;296(1):87–93.

   67. Larson AM, Polson J, Fontana RJ, et al. Acetaminophen-induced acute liver failure: Results of a United States multicenter, prospective study. Hepatology 2005;42(6):1364–1372.

   68. Laine L, White WB, Rostom A, Hochberg M. COX-2 selective inhibitors in the treatment of osteoarthritis. Semin Arthritis Rheum 2008;38(3):165–187.

   69. Grosser T. The pharmacology of selective inhibition of COX-2. Thromb Haemost 2006;96(4):393–400.

   70. McEvoy GK, ed. AHFS Drug Information®. Bethesda, MD: American Society of Health-System Pharmacists, Inc. ISBN 978-1-58528-247-0. ISSN 8756-6028. STAT!Ref Online Electronic Medical Library,, 2012.

   71. Lanas A. Nonsteroidal antiinflammatory drugs and cyclooxygenase inhibition in the gastrointestinal tract: A trip from peptic ulcer to colon cancer. Am J Med Sci 2009;338(2):96–106.

   72. Lazzaroni M, Porro GB. Management of NSAID-induced gastrointestinal toxicity: Focus on proton pump inhibitors. Drugs 2009;69(1):51–69.

   73. Rostom A, Moayyedi P, Hunt R, Canadian Association of Gastroenterology Consensus Group. Canadian Consensus Guidelines on long-term nonsteroidal anti-inflammatory drug therapy and the need for gastroprotection: Benefits versus risks. Aliment Pharmacol Ther 2009;29(5):481–496.

   74. Tannenbaum H, Bombardier C, Davis P, Russell AS, Third Canadian Consensus Conference Group. An evidence-based approach to prescribing nonsteroidal antiinflammatory drugs: Third Canadian Consensus Conference. J Rheumatol 2006;33(1):140–157.

   75. Chan FK, Hung LC, Suen BY, et al. Celecoxib versus diclofenac and omeprazole in reducing the risk of recurrent ulcer bleeding in patients with arthritis. N Engl J Med 2002;347(26):2104–2110.

   76. Bresalier RS, Sandler RS, Quan H, et al. Cardiovascular events associated with rofecoxib in a colorectal adenoma chemoprevention trial. N Engl J Med 2005;352(11):1092–1102.

   77. Solomon SD, McMurray JJ, Pfeffer MA, et al. Cardiovascular risk associated with celecoxib in a clinical trial for colorectal adenoma prevention. N Engl J Med 2005;352(11):1071–1080.

   78. Kearney PM, Baigent C, Godwin J, Halls H, Emberson JR, Patrono C. Do selective cyclooxygenase-2 inhibitors and traditional non-steroidal anti-inflammatory drugs increase the risk of atherothrombosis? Meta-analysis of randomised trials. BMJ 2006;332(7553):1302–1308.

   79. Olsen AM, Fosbol EL, Lindhardsen J, et al. Long-term cardiovascular risk of nonsteroidal anti-inflammatory drug use according to time passed after first-time myocardial infarction: A nationwide cohort study. Circulation 2012;126(16):1955–1963.

   80. Zhang J, Ding EL, Song Y. Adverse effects of cyclooxygenase 2 inhibitors on renal and arrhythmia events: Meta-analysis of randomized trials. JAMA 2006;296(13):1619–1632.

   81. Soni P, Shell B, Cawkwell G, Li C, Ma H. The hepatic safety and tolerability of the cyclooxygenase-2 selective NSAID celecoxib: Pooled analysis of 41 randomized controlled trials. Curr Med Res Opin 2009;25(8):1841–1851.

   82. Celebrex [package insert]. Pfizer, Inc., Mission, KS, 2012.|165564013|0&skwid=43700003360426290

   83. Derry S, Moore RA, Rabbie R. Topical NSAIDs for chronic musculoskeletal pain in adults. Cochrane Database Syst Rev 2012;9:CD007400.

   84. Bellamy N, Campbell J, Robinson V, Gee T, Bourne R, Wells G. Intraarticular corticosteroid for treatment of osteoarthritis of the knee. Cochrane Database Syst Rev 2006;(2):CD005328.

   85. Pyne D, Ioannou Y, Mootoo R, Bhanji A. Intra-articular steroids in knee osteoarthritis: A comparative study of triamcinolone hexacetonide and methylprednisolone acetate. Clin Rheumatol 2004;23(2):116–120.

   86. Habib GS, Bashir M, Jabbour A. Increased blood glucose levels following intra-articular injection of methylprednisolone acetate in patients with controlled diabetes and symptomatic osteoarthritis of the knee. Ann Rheum Dis 2008;67(12):1790–1791.

   87—capsaicin warning.

   88. Fraenkel L, Bogardus ST, Jr, Concato J, Wittink DR. Treatment options in knee osteoarthritis: The patient’s perspective. Arch Intern Med 2004;164(12):1299–1304.

   89. Cepeda MS, Camargo F, Zea C, Valencia L. Tramadol for osteoarthritis: A systematic review and metaanalysis. J Rheumatol 2007;34(3):543–555.

   90. Pergolizzi J, Boger RH, Budd K, et al. Opioids and the management of chronic severe pain in the elderly: Consensus statement of an international expert panel with focus on the six clinically most often used world health organization step III opioids (buprenorphine, fentanyl, hydromorphone, methadone, morphine, oxycodone). Pain Pract 2008;8(4):287–313.


   92. Chou R. 2009 Clinical Guidelines from the American Pain Society and the American Academy of pain medicine on the use of chronic opioid therapy in chronic noncancer pain: What are the key messages for clinical practice? Pol Arch Med Wewn 2009;119(7–8):469–477.

   93. Bellamy N, Campbell J, Robinson V, Gee T, Bourne R, Wells G. Visco supplementation for the treatment of osteoarthritis of the knee. Cochrane Database Syst Rev 2006;(2):CD005321.

   94. Samson DJ, Grant MD, Ratko TA, Bonnell CJ, Ziegler KM, Aronson N. Treatment of primary and secondary osteoarthritis of the knee. Evid Rep Technol Assess (Full Rep) 2007;(157):1–157.

   95. Hunter DJ. Osteoarthritis: Hyaluronic acid is not effective in symptomatic hip OA. Nat Rev Rheumatol 2009;5(7):359–360.

   96. Wandel S, Juni P, Tendal B, et al. Effects of glucosamine, chondroitin, or placebo in patients with osteoarthritis of hip or knee: Network meta-analysis. BMJ 2010;341:c4675.

   97. Clegg DO, Reda DJ, Harris CL, et al. Glucosamine, chondroitin sulfate, and the two in combination for painful knee osteoarthritis. N Engl J Med 2006;354(8):795–808.

   98. Berenbaum F. Targeted therapies in osteoarthritis: A systematic review of the trials on Best Pract Res Clin Rheumatol 2010;24(1):107–119.

   99. Brandt KD, Mazzuca SA, Katz BP, et al. Effects of doxycycline on progression of osteoarthritis: Results of a randomized, placebo-controlled, double-blind trial. Arthritis Rheum 2005;52(7):2015–2025.

  100. Kulkarni SK, Singh VP. Licofelone: The answer to unmet needs in osteoarthritis therapy? Curr Rheumatol Rep 2008;10(1):43–48.

  101. Lane NE, Schnitzer TJ, Birbara CA, et al. Tanezumab for the treatment of pain from osteoarthritis of the knee. N Engl J Med 2010;363(16):1521–1531.

  102. Kwon YD, Pittler MH, Ernst E. Acupuncture for peripheral joint osteoarthritis: A systematic review and meta-analysis. Rheumatology (Oxford) 2006;45(11):1331–1337.

  103. Scharf HP, Mansmann U, Streitberger K, et al. Acupuncture and knee osteoarthritis: A three-armed randomized trial. Ann Intern Med 2006;145(1):12–20.

  104. White AG, Birnbaum HG, Janagap C, Buteau S, Schein J. Direct and indirect costs of pain therapy for osteoarthritis in an insured population in the united states. J Occup Environ Med 2008;50(9):998–1005.

  105. Marra C, Cibere J, Grubisic M, et al. Pharmacist initiated intervention trial in osteoarthritis (PhIT-OA): A multidisciplinary intervention for knee osteoarthritis. Arthritis Care Res (Hoboken) 2012;64(12):1837–1845.

  106. Latimer N, Lord J, Grant RL, et al. Cost effectiveness of COX 2 selective inhibitors and traditional NSAIDs alone or in combination with a proton pump inhibitor for people with osteoarthritis. BMJ 2009;339:b2538.

  107. Hur C, Chan AT, Tramontano AC, Gazelle GS. Coxibs versus combination NSAID and PPI therapy for chronic pain: An exploration of the risks, benefits, and costs. Ann Pharmacother 2006;40(6):1052–1063.