Range of Motion and Stretching Exercises

ROM and stretching exercises should be part of every osteoarthritis patient’s daily routine. Beneficial effects reported include maintenance of function, decreased edema, stimulation of flexion-extension reflexes, and preparation of the limb for active exercise. Stretching exercises can restore or maintain ROM. Care should be taken with inflamed joints, because passive ROM exercise has been shown to increase joint inflammation. Although almost universally recommended, there is little scientific evidence for the use of stretching in the treatment of OA.

Strengthening Exercises

In knee arthritis, loss of strength and function occurs rapidly. A muscle can atrophy up to 30% in 1 week. A muscle at complete rest will lose strength at a rate of 3%/day.^81,82,198

Despite a wealth of published literature, consensus regarding optimum dosage, modality, and frequency of exercise for strengthening the quadriceps is lacking. In one case study,¹⁷⁷ isometric strengthening of the quadriceps muscles led to improvements in quadriceps torque, clinical status, and pain after walking. This program consisted of exercises performed three times weekly for 6 weeks with the knee flexed to 60 degrees. Other studies have demonstrated improvement in function of quadriceps-trained individuals, but most of these studies failed to compare the results to patients who rested. Three randomized controlled trials (RCTs)^70,76,133 in patients with knee osteoarthritis who underwent quadriceps strengthening with isometric, isotonic, or resistive exercises showed significant improvements in quadriceps strength, knee pain, and function when compared with controls. A recent meta-analysis of 10 RCTs has concluded that quadriceps strengthening has statistically significant treatment effects on pain and function in patients with osteoarthritis of the knee.²⁴²

Nevertheless, strengthening exercises must be used with caution. Exercises that use repetitive joint motion or require a full ROM may increase inflammation and pain, and thus fail to achieve muscle strengthening. Isometric contraction is less likely to increase joint pain or inflammation. Dynamic (repetitive) exercises are appropriate using isotonic or isokinetic muscle contraction after pain is controlled. Isokinetic exercises can be used for patients with ligamentous stability and no internal derangement. Deep knee bends, however, may increase intra-articular pressure and should be avoided.⁴¹

Aerobic Conditioning

In addition to weakness, patients with osteoarthritis often suffer from decreased cardiovascular endurance.¹²⁴ Aerobic exercise can increase the overall vitality, activity, and feeling of well-being in these patients. Suitable endurance exercises include cycling, swimming, and low-impact aerobics. High-impact loading activities, such as jogging, should probably be avoided.

Increased aerobic fitness not only improves the patients overall health, but specifically improves their arthritic symptoms. Kovar and coworkers,¹⁵² in an 8-week supervised fitness walking program in 102 patients with knee osteoarthritis, found improvements in 6-minute walking distance and reductions in pain and the use of medications in the exercise group when compared with controls. In another study, 12 weeks of aerobic walking or aquatic exercises improved overall exercise capacity, with aerobic gains maintained at 9 months. In the Fitness, Arthritis and Seniors Trial,⁷¹ 439 subjects with radiographically confirmed osteoarthritis of the knee, pain, and disability were randomized to a program of aerobic exercise, resistance exercise, or health education. In this 18-month trial, those in the exercise groups showed improvements in tests of physical performance (climbing and descending stairs, lifting and carrying 10 pounds) compared with the education group. Pain and disability self-reported scores improved in the exercise groups.

Roddy and colleagues have published a systematic review of 13 RCTs. They reported a significant treatment effect for aerobic conditioning and quadriceps strengthening in patients with osteoarthritis of the knee.²⁴²

Taping

Appliances that alter the biomechanics of the knee joint may be helpful. Cushnaghan and associates⁵⁰ have found that taping the patella medially is effective in reducing knee pain in patients with patellofemoral arthritis. In this randomized, single-blind, crossover trial with 14 subjects, medial taping was superior to lateral or neutral taping for pain scores, symptomatology, and patient preference. Others have used taping before quadriceps exercises for chondromalacia patellae.¹⁸⁶

Knee Bracing and Orthotics

The usefulness of bracing for the treatment of knee osteoarthritis has been controversial.^{88,146,250,266} Most clinical and biomechanical studies have shown little or no benefit from these devices. The primary goal of knee brace designs is to assist in the restoration of normal mechanical stability. The first step in fitting for a brace is to define abnormal motion that the brace should control.

A Swedish knee cage or a hinged knee brace may provide support in limiting extension and may help decrease pain.²⁵⁰ There are a number of three-point pressure braces to control medial or lateral instability.²⁶⁶ In selected patients, these devices can be effective. In one study, Kirkley and associates¹⁴⁶ found that valgus-producing functional knee braces were much more effective for the treatment of medial compartment osteoarthritis of the knee than a simple neoprene sleeve. Furthermore, quality of life (Western Ontario and McMaster Universities Osteoarthritis Index [WOMAC]) scores of both braced groups exceeded those of a control group receiving standard medical treatment in a prospective, parallel group, randomized clinical trial.

Immobilization of the knee during periods of increased pain or inflammation may be useful. This may be done with a knee immobilizer or posterior splint. These support the knee in extension and permit relaxation of the flexor muscles. An elastic bandage may control knee swelling but is actually not a knee orthosis.

Assistive Devices: Cane or Walker

The cane can successfully unload the knee joint.²³ Assistive devices to unload the knee joint are most effectively used on the side opposite the pathologic condition. The use of a crutch or cane will reduce the joint load on the opposite limb by about 50%. A quadriceps cane can be used instead of a straight cane when balance is a problem.

Heat Modalities

Therapeutic heat can be applied superficially or to a deep location. Heat is usually applied at temperatures of 41° C to 45° C. Superficial heat is capable of elevating the soft tissue temperatures by 3° C at a depth of 1 cm without penetrating deeper depths and thus penetrates the knee joint.¹²⁸ Some studies have demonstrated that the threshold for pain can be raised in humans as well as in animals by the application of superficial or deep heat.¹⁶³ The effect is produced by analgesia of free nerve endings (peripheral nerves and gamma fibers of muscle spindles) and muscle relaxation is produced.¹⁶⁴ Local heat may also relieve pain by acting on sensory afferents and closing the pain gate, or increasing local blood flow and thus washing out pain-inducing metabolites and inflammatory mediators produced in osteoarthritis.

The four general methods that produce superficial heat include diathermy (shortwaves) microwaves, ultrasound, radiation (infrared), conduction (heating pad, water bottle), and convection (sauna, steam room).^{57,66,211,234} Moist heat produces a greater temperature elevation than dry heat and may be preferable for clinical applications.¹²⁸ For all these modalities, care must be exercised to avoid burns, especially with uneven application. A towel-wrapped hot water bottle, gel-filled hot pack, or thermostatically controlled electric heating pad provides a simple method for the patient to benefit from superficial heat application at home.

Interferential Therapy

Interferential therapy uses two medium-frequency (approximately 4 KHz) alternating currents applied to the skin through suction cups or adhesive padding. The resultant current has a low frequency that is the difference between the two original frequencies applied. This current is usually applied for about 15 minutes for the knee and is experienced as a prickling sensation. Various pain-relieving mechanisms, which block nonmyelinated nociceptive fibers, activate delta A and C fibers, releasing encephalins and endorphins, or activatef the opioid system, have been proposed for the pain relief that this technique provides.²³⁴

Transcutaneous Electrical Neuromuscular Stimulation

Transcutaneous electrical nerve stimulation (TENS) delivers short pulse width (50 to 250 microsecond), low-frequency waves (2 to 150 Hz) that are used specifically for pain relief.¹³⁴ As with interferometry, a prickly sensation is produced. Carbon-rubber electrodes with a coupling gel on the skin or with self-adhesive electrodes are used to deliver pulses for 30 to 60 minutes once or twice daily. The finding that large-diameter, cutaneous nerve fibers are preferentially stimulated by TENS is thought to account for its efficacy. These fibers inhibit the transmission of painful stimuli to the spinal cord. Double-blind studies using TENS for the control of pain in osteoarthritis (OA) of the knee have yielded conflicting results. One study concluded that it provides no greater benefit than placebo whereas the other study demonstrated improvement over placebo.¹¹⁵ A recent meta-analysis was inconclusive; it found only small, poorly constructed studies.²⁵¹ More research is needed to demonstrate the effectiveness of this modality and define optimum parameters for its use.

Magnetism, Electrical Stimulation, and Low-Intensity Pulsed Ultrasound

Magnetism, electrical stimulation, and low-intensity pulsed ultrasound use low-energy fields to achieve their effects. The device that delivers a pulsed electrical signal has been approved by the U.S. Food and Drug Administration (FDA) for the treatment of OA of the knee and rheumatoid arthriis (RA) of the hand. Zizic and associates³¹⁰ have published a short-term clinical study comparing patients with OA of the knee treated with this device versus those treated with placebo. The treatment group had statistically significant improvement in pain and function scores over the 4-week treatment period. There is even some indication that these modalities have a disease-modifying effect. Low-intensity pulsed ultrasound was recently shown to yield more hyaline-like cartilage in a rabbit model when compared with non-treated controls in a study by Cook and coworkers.⁴⁸ Lippiello and colleagues,¹⁶⁸ using this device, also reported improved hyaline cartilage repair in a ulcerative and bone defect model in rabbits when the animals were treated with electromagnetic stimulation. In a recent Cochrane database review of TENS for OA of the knee, the efficacy could not be confirmed, given the small number of clinical studies of questionable quality.²⁵¹

Nonselective Nonsteroidal Anti-inflammatory Drugs

Inhibition of prostaglandin synthesis has detrimental effects. Prostacyclin (PGI2) and prostaglandin E2 (PGE2) are both vasodilators important in maintaining renal perfusion during hypovolemia. Prostaglandin inhibition leads to sodium retention in the kidney, which may worsen congestive heart failure. A number of medical conditions depend on renal prostaglandins to maintain renal profusion, including congestive heart failure, cirrhosis, certain forms of hypertension, and dehydration. In these patients, exposure to NSAIDs will lead to a decline in renal function, even if creatinine clearance was normal before treatment.^46,202 This decline is usually reversible. NSAIDs may increase the overall risk of chronic renal failure.²²⁵ Acute interstitial nephritis has also been noted with most of the NSAIDs, but is seen most commonly with fenoprofen.^37,176 sulindac,^299,300 and nabumatone⁵ are purported to be less likely to cause deterioration in renal function.³⁸

The most common toxic side effect of traditional NSAIDs occurs in the GI tract. It has been estimated that 15% to 35% of all peptic ulcer complications are attributable to these drugs.^{113,114,116,120,156}

Morbidity includes 107,000 hospitalizations and 16,500 deaths annually in the United States alone.²⁶⁴ Upper GI symptoms caused by NSAIDs include dyspepsia, ulceration, hemorrhage, perforation, and death.^94,99,265 There is an estimated three- to fivefold increased relative risk of GI bleeding from the use of NSAIDs.¹¹⁹ Risk factors for developing a bleeding complication with the use of these medications include a history of prior peptic ulcer disease, concomitant use of corticosteroids or anticoagulants, and poor general health.^{92,98,261,263} Toxicity from NSAIDs is additive, so the use of more than one NSAID at a time is contraindicated.^40,189 The concomitant use of NSAIDs in patients who are also taking systemic corticosteroids should be avoided, if possible, because the incidence of bleeding complications and mortality is significantly elevated.

Selective COX-2 Inhibition

COX-2 inhibitors, also called coxibs, reduce GI adverse events and complications.^56,67,254 Two large, prospective, randomized outcome studies were performed for celecoxib and rofecoxib.^25,262 Celecoxib was compared with diclofenac and ibuprofen and rofecoxib was compared with naproxen. The risk of symptomatic ulcers or ulcer complications was lower with the selective COX-2 inhibitors.

The selectivity of COX 2 inhibitors, although beneficial to the GI mucosa, may lead to problems with thrombosis as well as salt and fluid imbalance.³⁰⁵ The incidence of acute myocardial infarction was significantly higher for rofecoxib than naproxen in this study. The celecoxib did not demonstrate increased risk or cardiac events compared with ibuprofen or diclofenac. An independent reanalysis of the celecoxib data confirmed that there was no increased risk of myocardial infarction with celecoxib.³⁰¹ Two large chemoprevention trials for adenomatous polyps, one with celecoxib and one with rofecoxib, were discontinued after demonstrating increased cardiovascular risk.^75,187 The dosages of the medications in these studies were far greater than those used to treat arthritis. As a result of its increased cardiac risk profile, rofecoxib was removed from the market.

The exact mechanism for increased cardiac risk has not been elucidated. In addition to vasoconstriction, which can be associated with COX-2 inhibition, a number of other factors may cause increased cardiac risk. The COX-2 enzyme expression is found in endothelial cells in response to injury.¹³⁹ It is also found in atheromatous plaques and may play a role in decreasing vascular inflammation.²⁵⁸ Furthermore, there is no COX-2 expression on platelets. COX-2 inhibitors, unlike their nonselective counterparts, do not block formation of thromboxane, which plays an important role in platelet aggregation and vasoconstriction. This may explain the apparent increased cardiovascular risk of coxibs over traditional NSAIDs. A recent trial of anti-inflammatory drugs in patients with Alzheimer’s disease indicated that naprosyn, 220 mg, had an increased risk of cardiovascular risk that was higher than celebrex, 200 mg, in older patients.¹ In a report of population-based National Health Insurance data from Taiwan, four NSAIDs were examined.¹³⁰ In patients 18 years of age or older, there was no difference in the risk of serious long-term events in patients treated with etodolac, ibuprofen, naproxen, nambutone, or celebrex for an 180-day period. In this study, a previous history of cardiovascular disease was the greatest predictor of risk.

Hypertensive effects of COXIBs enzymes seem equal to those of nonselective NSAIDs. It appears that COX-2 enzyme is responsible for prostaglandin production,which is important for fluid balance. Blocking its production affects fluid retention, which can result in hypertension.¹³⁹ The COX-2 enzyme is also responsible for the production of PGI2, which is a vasodilator.⁸³ A number of products of the COX-1 enzyme have vasoconstrictive effects. Therefore, selective inhibition of COX-2 favors vasoconstriction, which plays a role in hypertension and heart disease.

Nonsteroidal Anti-Inflammatory Drugs versus Analgesics

Although a number of studies have shown that NSAIDs are more effective than placebo, evidence that they are more effective than simple analgesics (such as acetaminophen) in osteoarthritis is not consistent. Several studies have demonstrated a 10% to 20% improvement in scores of pain and stiffness when NSAIDs were compared with placebo.^17,20,280 In a large comparative trial, slightly less than 50% of patients exhibited a favorable response to treatment. No difference in response rates among drugs was noted.¹⁷⁰

A 4-week randomized, prospective, blinded study comparing an anti-inflammatory dose of ibuprofen, an analgesic dose of ibuprofen, and acetaminophen failed to show any significant difference among the three treatment groups.²⁸ Another study comparing ketaprofen with dextropropoxyphen-acetaminophen (a non-narcotic analgesic) failed to show any difference between the regimens.⁶³ Consistent with these findings have been several studies in which ibuprofen given in an analgesic dose (1200 mg) was equivalent to several other NSAIDs for the treatment of osteoarthritic joint pain.^{32,44,55,197,249} Given the evidence, one might consider that the efficacy of NSAIDs lies in their analgesic effect rather than in the anti-inflammatory one. Schumacher and associates have shown that high-dose ibuprofen is superior to acetaminophen in patients with knee arthritis and a synovial effusion. Therefore, there may be a subset of patients with OA and an inflammatory component who benefit additionally from NSAIDs.²⁵⁹

A recent study has compared an extended-release acetaminophen, 1300 mg three times daily, to rofecoxib 12.5 and 25 mg once daily.²⁵⁷ Acetaminophen was noninferior to rofecixib at the 12.5-mg dose. Some studies have shown some potential benefit of NSAIDs over acetaminophen. In a 6-week trial comparing acetaminophen with diclofenac and misoprostol, the latter had a statistically higher response to treatment for OA of the hip and knee.²²⁷

NSAIDs are considered more effective for symptomatic treatment than acetaminophen. In a meta-analysis of 10 randomized control trials including 1712 patients,³⁰⁷ NSAIDs were more effective for pain relief and had a better clinical response rate than acetaminophen. Twice as many patients preferred NSAIDs to acetaminophen. However, NSAIDs were more commonly associated with GI side effects, including nausea, vomiting, distress, abdominal pain, and diarrhea.

Another meta-analysis of 23 randomized, placebo-controlled studies has demonstrated that NSAIDs are effective in pain reduction in osteoarthritic knees over short periods.²² This analysis included selective COX-2 inhibitors. Although the reduction in pain with NSAIDs is overall greater than that seen with acetaminophen, the difference may not be enough for clinical significance.³⁰⁷ NSAIDs are associated with higher rates of GI discomfort than Tylenol.²⁷⁸

There is some evidence to suggest that NSAIDs interfere with synovial blood flow or repair of microfractures of subchondral bone.^270,271 Furthermore, various animal models have shown that the most commonly used NSAID, aspirin, inhibits proteoglycan synthesis and leads to enhanced cartilage destruction.^{29,30,213–215} Early clinical studies implied that NSAIDs were associated with more rapid degeneration of the osteoarthritic joint and quicker presentation for surgery. A retrospective analysis of radiographic progression of osteoarthritis of the hip has concluded that indomethacin is associated with greater joint destruction than that seen in control patients.²⁴⁶ In another study, patients with advanced hip OA awaiting arthroplasty were treated with indomethacin or azapropazone.²³⁶ Azapropazone is a nonselective NSAID available for use in Britain. It was concluded that the indomethacin group showed more rapid radiographic deterioration and presented to surgery earlier then the azapropazone group. However, the results of this study have been called into question,⁶² and a double-blind study of indomethacin versus placebo showed no increase in disease progression.¹²⁵

Possible Chondroprotective Action of Nonsteroidal Anti-Inflammatory Drugs

The traditional view that osteoarthritis is an inevitably progressive disease and results from wear and tear of the cartilage has been replaced by an understanding of the biochemical and biomechanical factors in the cause and progression of the disease. Because cartilage is continuously undergoing degradation and renewal, it appears logical to design a medication that promotes anabolic activity of cartilage while inhibiting its degradation. The evidence for these beneficial activities of NSAIDs is mixed.¹²³

A chondroprotective effect of NSAIDs has been postulated.⁶¹ Proposed mechanisms include improved biomechanics by decreasing arthralgia and inhibition of cartilage catabolism. Cartilage matrix proteoglycans are degraded by enzymes such as metalloproteases and serine proteases. Some NSAIDs are effective inhibitors of these enzymes.^34,101,165 Release of oxygen free radicals and other inflammatory mediators may also be suppressed by NSAIDs.^122,221,224 Other NSAIDs may actually stimulate glycosaminoglycan production, as indicated by increased sulfate incorporation.

The net effect of NSAIDs on cartilage remains to be determined and may vary among NSAIDs.²¹⁴ The effects of COX-2 inhibitors on articular cartilage have not been extensively studied. Celecoxib and diclofenac on human chondrocyte metabolism were compared in an in vitro model. Celecoxib increased the synthesis of hyaluronan and proteoglycans in these explanted cells, whereas diclofenac did not have such an effect.⁶⁹

Other studies have also demonstrated a potential chondroprotective effect of COX-2 inhibitors in vitro. Human articular cartilage cells exposed to celecoxib in culture have demonstrated increased proteoglycan synthesis and decreased proteoglycan release.¹⁸⁰

Future Medications

COX-inhibiting nitric oxide donators (CINODs) are a recently developed group of analgesic and anti-inflammatory medications. It is theorized that the addition of the nitrogen oxide (NO) will counteract some of the known complications seen with COX-inhibitors, specifically elevated blood pressure and gastrointestinal upset.²⁹⁵ The release of NO causes vasodilation of blood vessels, which decreases systemic blood pressure and decreases platelet aggregation. The first drug in this class is AZD3582 or naproxcinod, a combination of naproxen and nitric oxide.

White and coworkers³⁰² have recently conducted a study to compare the effects of naproxen alone versus naproxcinod on blood pressure. The treatment protocols for the four comparison groups were as follows: (1) 750 mg naproxcinod twice daily; (2) 375 mg naproxcinod twice daily; (3) naproxen 500 mg twice daily; and (4) placebo twice daily. The authors found that neither dose of naproxcinod resulted in increased blood pressure compared with the naproxen group. Also, patients who had a known diagnosis of hypertension prior to beginning the study who were treated with naproxen alone had an elevated blood pressure, 6.5 mm Hg higher than those hypertensive patients in the 500-mg naproxcinod treatment arm.

Karlsson and associates,¹⁴⁰ in a phase 2, randomized, double-blind study, found the most efficacious dose of naproxcinod to be 750 mg twice daily. Included in the study were patients treated with 25 mg daily of the COX-II inhibitor rofecoxib. This study found that no statistically significant difference in WOMAC pain score for those treated with naproxcinod, 750 mg twice daily; naproxcinod, 1125 mg twice daily; and rofecoxib, 25 mg daily. However, patients on these regimens had better pain relief than those taking a once-daily dose of naproxcinod, 750 mg, or placebo. The authors also reported decreased systolic blood pressure in the cohort being treated with rofecoxib.

Although these early results are promising, long-term follow-up of the CINOD group of drugs must be conducted to identify possible idiosyncratic side effects that have not been as yet identified.

Summary

If nonpharmacologic measures fail, acetaminophen should probably be tried as a first-line measure. In the subset of patients in whom analgesic therapy is not effective, a low dose (less than 1200 mg/day) of ibuprofen often provides effective therapy. Low-dose ibuprofen is not anti-inflammatory and is very inexpensive. If the ibuprofen is not effective in an analgesic dose, anti-inflammatory doses of ibuprofen or other nonselective NSAIDs may be tried. As a first-line pharmacologic treatment, the AAOS clinical practice guideline recommends acetaminophen or NSAIDs.²

In patients with GI risk, acetaminophen, COX-2 inhibitors, or nonselective NSAIDs with a gastroprotective agent such as a protein pump inhibitor may be used.^2,309 All NSAIDs should be used cautiously in patients with hypertension, a history of cardiac or renal disease, and in older adult. Scheiman and Fendrick²⁵⁵ have provided an algorithm to help determine the appropriate NSAID based on gastrointestinal and cardiovascular risk (Table 92-1).

Table 92-1 Clinicians Guide to Anti-Inflammatory Therapy

Once relief is achieved with NSAIDs, periodic withdrawal of therapy with substitution of a simple analgesic is prudent, especially in the older population. NSAIDs should be avoided in high-risk patients, such as those with a history of ulcer disease, those on concurrent oral corticosteroids, history of gastrointestinal bleeding, congestive heart failure, or renal insufficiency. Nonacetylated salicylates (Arthropan), salsalate (Disalcid), choline magnesium trisalicylate (Trilisate), or renal-sparing NSAIDs such as sulindac may have a role in the treatment of these patients.