A 57-Year-Old Man With Osteoarthritis of the Knee

DR SHIP: Mr V is a 57-year-old man who has had knee pain for at least 30 years. He lives in Boston with his wife, with whom he owns and runs a business. He has managed care insurance and sees his primary care physician, Dr P, at Beth Israel Deaconess Medical Center.

Mr V dates the onset of his left knee problems to a soccer injury at age 26. He was diagnosed with rupture of 2 ligaments and underwent a course of physical therapy. His pain diminished over time, but he was aware that his left knee was never again at "100%." He continued his active athletic life, which included soccer, bicycle racing, and running at least 3 times per week. About 15 years ago, however, pain in his left knee recurred and became severe enough that he stopped running and moved largely to long-distance cycling. Thirteen years ago he underwent an arthroscopic debridement. This improved his level of function and decreased his pain for approximately 2 years, but symptoms subsequently recurred. He has since had several courses of physical therapy but has found the recommended exercises to be so painful that he could not complete them.

Despite persistent, often severe, knee pain, Mr V remains a very active long-distance ("ultra marathon") bicyclist, cycling in qualifier events and races of more than 700 miles (1120 km). He estimates his annual cycling distance to be about 7000 miles (11 200 km), a number that he views as a marked reduction from his previous distance of between 10 000 and 12 000 miles (16 000-19 200 km) annually. In addition, he continues to lift weights and work out regularly. His knee pain has progressed, however, so that he often has severe pain after standing for long periods as he needs to during work, and he is unable to garden because of pain brought on by bending his left knee. He was prescribed naproxen, 1000 mg/d, but finds that 500 mg/d is effective at "helping" with the pain, and he is hesitant to take the higher dose.

Other than knee pain, Mr V is very healthy. He has no other medical problems and has had no other surgical procedures. He follows a vegetarian diet and neither smokes nor drinks. In addition to the naproxen, he takes garlic pills. He is not allergic to any medications. There is no family history of osteoarthritis.

On physical examination, Mr V was 5'10" tall and weighed 155 lb. His left knee appeared slightly larger than his right but had no erythema or warmth. When standing, he had varus deformities of both knees and a slightly antalgic gait. Range of motion on the left knee was 5° to 120° compared with 5° to 140° on the right. Collateral and cruciate ligaments were stable. Crepitus was noted. There was a moderate effusion in the left knee and a palpable semi-membranous bursa. There was no distal neurovascular compromise in either knee. Hip range of motion was full and pain-free bilaterally.

Radiographs of the left knee in 1999 revealed marked tricompartmental osteoarthrosis with prominent osteophyte formation and severe joint space narrowing. The medial joint space was the most narrowed compartment. There was subchondral sclerosis, most prominently also in the medial hemi-joint space, and lateral subluxation and varus angulation were noted. Significant subchondral cystic changes were identified involving the tibial plateau centrally as well as the medial femoral condyle, thought to represent geode formation from osteoarthrosis.

Total knee replacement has been offered to Mr V, but different orthopedists have varied in their recommendations regarding the timing of the surgery. Mr V wonders whether and when he should plan to undergo knee replacement and how long he can expect the prosthesis to last.

I hurt my knee playing soccer some 30 years ago, and in the beginning it was not a big deal. For the longest time I kept playing and doing other sports such as cycling. But it became so bad that I couldn't even run anymore. So I dedicated all my time to cycling.

Cycling can cause pain if I don't use the right gear—if I push myself beyond the point that I know I shouldn't, and often I do. Knee pain mainly happens when I stand too long and that happens at work a lot; and walking long distances, not wearing proper shoes, walking on hard surfaces, and stuff like that. My left leg is totally bowed. I know I have a limp that I don't even notice anymore.

I saw a doctor recently and immediately he wanted to put me in physical therapy and he said that it's not necessary to have a knee replacement yet. He thought I was way too young. I'm 57 years old. The next doctor I saw said, "Well, that's his opinion. I don't agree with it. Seeing your condition, I think it's a waste of time to spend this much time in physical therapy."

Until my final decision, I would like to talk to some more specialists and see what other people have to say about it. My main question is should I continue with my long-distance cycling or should I stop? The last doctor said that he wouldn't stop me from riding but he wouldn't guarantee that the success of the surgery would be certain if I keep riding the way I ride today.

Mr V is certainly an interesting contrast because on the one hand, he has very severe pain and very severe arthritis radiographically, and it clearly limits him. On the other hand, he is a testament to what the human spirit can endure because he still races incredibly long distances.

I think my first reaction is always to try to ward off the "evil demon" of surgery for as long as possible, particularly for somebody who at first glance seems to be doing well. He exemplifies all the things that we would normally tell people to do to manage it medically. He exercises regularly; he's trim.

The questions about the timing of the surgery are always the most difficult. It seems like many people struggle with this for a long time before they get their surgery and endure what seems to be unnecessary suffering. Is there any role for doing less than entire surgery? That seems like sort of a cop-out, but if surgeons could help patients avoid some of the problems with knee replacements that would be great.

What are the epidemiology and natural history of osteoarthritis (OA) affecting the knee? At what age does OA usually become symptomatic? What are the usual presenting symptoms? What specific aspects of the physical examination are helpful or diagnostic? What other disease processes should be considered in a middle-aged patient with knee pain? What imaging is recommended? What are the treatment options and what is the evidence for their success? When is knee replacement recommended? What kind of recovery is necessary? What are the potential complications? What is the anticipated life span of the replacement? What do you recommend for Mr V?

Dr Lonner: Mr V is an avid cyclist with severe unilateral knee pain. He has been treated with a variety of interventions including physical therapy, nonsteroidal anti-inflammatory medications, and activity modification. It is interesting that Mr V has less pain with long-distance cycling than prolonged standing or walking. This suggests that his arthritis may be predominantly tibiofemoral, with less patellofemoral involvement. The progressive varus bowing of his knee is indicative of medial compartment arthritis. He now seeks further intervention in light of his progressive difficulty performing the important responsibilities of his profession and routine daily activities, despite appropriate interventions.

As the most prevalent form of arthritis, OA affects 16 million to 40 million people in the United States, accounting for 43% of arthritis.^{1 - 2} While primarily affecting elderly individuals, OA also is a common cause of impairment and disability in the middle-aged population.³ The knee is the most commonly affected joint, with more than 4 million new patient visits annually in the United States for knee pain.³ While radiographic evidence of OA is common by the age of 40, clinically significant OA of the knee is less common until the age of 60, after which approximately 25% of the general population has pain and disability related to OA.⁴

Left untreated, unicompartmental arthritis will progress over 10 to 20 years in 50% to 60% of malaligned limbs.^{5 - 6} However, the prevalence and severity of OA of the knee tend to remain constant during the seventh through ninth decades.⁷

Many patients have no identifiable cause for OA. Others may have secondary causes including chronic instability, prior meniscectomy, obesity, injury, infection, crystalline disease, malalignment, or other anatomic derangements.^{8 - 9} A genetic predisposition is also likely.² Unlike excessive and repetitive loading, moderate and even occasional strenuous regular activity does not likely cause or accelerate osteoarthrosis in joints with "normal" articular surfaces, alignment, stability, innervation, and muscle control.^{10 - 11} Patients with derangement of the knee or OA may be advised to be more careful, restricting themselves to moderate activity levels.

Like Mr V, patients with osteoarthrosis of the knee commonly present with progressive, insidious discomfort hastened by activities such as walking, climbing, and kneeling. Patients with a high pain tolerance may fatigue with activity, despite a paucity of pain, representing a "symptomatic equivalent." Progressive joint stiffness and contracture may develop from osteophyte formation or synovitis and capsular scarring.² In severe cases, the pain may interfere with normal sleep patterns. Warmth, swelling, and crepitus are common. Patients may also describe mechanical symptoms such as the knee buckling or giving way. Arthritis often produces little or no pain at rest, which can help distinguish it from other conditions such as tumor or infection.

Physical examination should first include an inspection of gait to detect the presence of a limp, deformity of the knee, or instability. Pain referred from the back should be ruled out. A painful active straight leg raise and restricted ipsilateral hip motion that reproduces knee pain suggests hip pathology as a source of referred knee pain and should prompt radiographic study of the hip. The knees should also be examined for muscle atrophy and motion.

The 3 compartments of the knee should be assessed individually to localize tenderness and crepitation. Joint line tenderness may be observed in the presence of arthritis or meniscal tears, but provocative maneuvers, such as the McMurray test, can help diagnose the latter. Osteoarthritis and degenerative meniscus tears often coexist and can be difficult to sort out on physical examination alone. Ligamentous integrity, as well as patellar excursion, tilt, and tracking, all of which can cause knee pain, should be evaluated.

Periarticular palpation is performed to identify other potential sources of knee pain, including osteonecrosis, in which case tenderness may be localized to the proximal tibial or distal femoral metaphyses; pes anserinus bursitis, in which case tenderness and swelling would be focally located over the pes anserinus bursa (at the medial proximal tibial metaphysis); fatigue fracture, which often presents with metaphyseal tenderness; or tendonitis, which will present with tenderness over the affected tendon fibers.

The differential diagnosis of knee pain in patients aged 40 to 60 years includes the conditions in Table 1. Meticulous clinical examination can help delineate the cause of knee pain.

Weight-bearing radiographs should be obtained, including standing anteroposterior, lateral, and patellar views. Standing midflexion posteroanterior radiographs can help identify chondral degeneration isolated to the posterior half of the tibiofemoral compartments. Except in the rare patient with a suspected fracture or impending fracture, all knee films should be taken with the patient standing because supine radiographs may underestimate the true extent of chondral loss. Mr V's radiographs are shown in Figure 1. Typical weight-bearing radiographs and patellar views will show the extent of articular cartilage degeneration, the presence of hypertrophic osteophytes, and cystic formations such as that seen in Mr V. Additionally, they will identify the presence of tibiofemoral subluxation and the possibility of patellar tilt or subluxation. The lateral radiograph will show most effectively the presence of anterior or posterior osteophytes as well as the vertical position of the patella relative to the joint line.

Magnetic resonance imaging (MRI) is commonly overused in patients with arthritis. Clinical experience has demonstrated that routine MRI for patients with midstage osteoarthrosis is neither advisable nor cost-efficient. Meniscal and anterior cruciate ligament degenerative changes are common in middle-aged and elderly patients with osteoarthrosis and should not be misinterpreted as being the source of pain.¹² However, MRI may be helpful in ruling out acute sources of knee pain, such as meniscal tears, osteonecrosis, or insufficiency fracture.¹³ Although specialized MRI methods have been developed for assessing focal chondral defects, these imaging methods generally are of little use in patients with more advanced arthrosis.¹⁴

Before considering surgery, patients should be treated with nonoperative interventions, including activity modification, weight loss (if appropriate), physical therapy, and judicious use of medication, many of which nonsurgical physicians can manage prior to referral to an orthopedist. Mr V seems to have exhausted many of the nonsurgical options available. Referral to an orthopedic surgeon is appropriate when the nonsurgical interventions are ineffectual.

Physical Therapy and Exercise. Mr V exercises frequently and reports that his long-distance cycling does not cause his pain to escalate, supporting the notion that for some, a carefully guided exercise program may help reduce the pain of OA. However, exercise done excessively or inappropriately may worsen arthritic symptoms. Patients with arthritis cannot be treated like those without: exercise programs should be individually tailored according to the patient's age, severity of arthritis, and generalized conditioning.¹⁵ Nonimpact aerobic workouts and appropriate exercises can enhance cardiovascular conditioning, facilitate weight reduction, increase muscle strength and endurance, maintain joint motion, reduce pain and enhance proprioception, and encourage independence and functionability.^{15 - 16} Mr V has tried physical therapy, but it has not provided long-term relief; alternatively, cycling has been therapeutic. Active or active-assisted range of motion exercises and stretching help maintain and even increase joint motion.^{15 - 16} Use of excessive weights, cartilage loading, or strengthening in an inappropriate arc may hasten the symptoms of arthritis.¹⁶ However, isometric exercises, in which muscle contraction is performed without joint motion, allows strengthening of the muscles with limited increase in intra-articular pressures.¹⁶ In patients with patellofemoral arthritis, strengthening through the terminal 30° of extension may allow isotonic or isokinetic exercises to be performed with a lower risk of symptom exacerbation.¹⁶ The benefits of aquatic therapy in improving aerobic capacity, physical activity performance, and depression indices after as little as 12 weeks support its use.¹⁷

Heel Wedges, Bracing, and Canes. Some patients with medial compartment arthrosis may benefit from a 5° lateral orthotic wedge or a valgus unloader brace, which may reduce external varus moments and medial compartment loads. Clinical studies have reported pain reduction and enhanced function 6 to 12 months after initiating treatment, but the acute and mid-term benefits vary and tend to decline with advanced arthritis.^{18 - 21} Use of a cane can reduce pain and improve function in patients with knee arthritis by improving balance and reducing joint reaction forces.²² In varus arthritis of the knee, using the cane in the ipsilateral hand helps prevent shifting of the center of force medially; in valgus arthritis, the cane in the contralateral hand may shift the center force medially, away from the affected area of the knee.²² Controlling the cane to impact the ground simultaneously with heel strike may decrease the peak vertical reaction forces by 25% to 30% at heel strike and 20% to 25% at midstance of the gait cycle.²³ Finally, shoes with soft, impact-absorbing heels and soles may reduce pain with ambulation.

Acetaminophen. Since 1995, acetaminophen has been recommended as the first-line agent for the treatment of OA by the American College of Rheumatology.²⁴ Given orally in divided doses of 4000 mg/d, acetaminophen is superior to placebo and as beneficial as ibuprofen and naproxen in reducing the pain of OA and enhancing joint function.²⁵ Patients should be monitored carefully for liver toxicity when taking acetaminophen long-term. Recently, gastric adverse effects have been reported. In a case study, the risk of gastrointestinal adverse effects was elevated, particularly in patients receiving doses higher than 3250 mg/d compared with those receiving 650 mg/d or less.²⁶

Nonsteroidal Anti-inflammatory Drugs. Mr V has tried taking nonsteroidal anti-inflammatory drugs (NSAIDs), one of the most commonly prescribed classes of medications. NSAIDS tend to be most effective for early arthritis and their pain relief is often only partial,²⁷ as seen in Mr V. It should be stressed that proper regular dosing of NSAIDs is far more efficacious than sporadic dosing; therefore, NSAIDS should be prescribed in regular dosing intervals for patients with chronic arthritis. The primary adverse effect of NSAIDs is gastrointestinal toxicity, with gastroduodenal ulceration observed endoscopically in as many as 15% to 30% of patients using NSAIDs.²⁸ Erosive ulcers have been noted after as little as 7 days of standard NSAID use at commonly prescribed dosages. This risk may be decreased with cyclooxygenase 2 (COX-2) selective inhibitors, which likely provide the same therapeutic anti-inflammatory and analgesic benefits as traditional nonselective NSAIDs, but with decreased toxicity on nontarget tissues such as the gastric mucosa,^{29 - 30} although at 15 to 20 times the cost of nonselective NSAIDS.²⁴ COX-2 inhibitors have a reported risk of cardiovascular toxicity of less than 1%.^{29 - 30} Renal damage is another potential complication of chronic NSAID use. Given the potential deleterious effects, ongoing treatment with NSAIDs should be withheld until other medications such as acetaminophen or nonpharmacologic modalities, such as weight loss or activity modification, have been tried.

Nutritional Supplements. Oral glucosamine and chondroitin sulfate are nutritional supplements, which are not regulated by the US Food and Drug Administration (FDA). Few formal, scientifically rigid studies have been performed to determine the efficacy or long-term safety of these supplements. Nonetheless, they may have value for the patient with OA of the knee.

Glucosamine may have an anti-inflammatory effect independent of prostaglandin inhibition because glucosamine does not inhibit cyclooxygenase.³¹ The most common dosage is 1500 mg/d of glucosamine sulfate. Small trials suggest short-term efficacy and safety of oral glucosamine.^{32 - 33} In animal studies, chondroitin sulfate has been shown to be chondroprotective and anti-inflammatory.^{34 - 35} Chondroitin sulfate is often given in doses ranging from 800 to 1200 mg/d. Compared with placebo, chondroitin sulfate may significantly decrease pain, reduce concomitant NSAID use, and decrease the amount of joint space narrowing in knee OA.³⁶ One 3-year, prospective, double-blind, placebo-controlled study assessing the long-term effects of glucosamine sulfate on OA progression found that 1500 mg/d of oral glucosamine sulfate reduced the tendency of further joint space destruction, limited pain, and enhanced function compared with placebo.³⁷ Although these supplements appear to be efficacious and safe for short-term use in OA, long-term risk analysis and further study are necessary. The FDA regulations regarding processing standards do not apply to manufacturers of these nutritional supplements and therefore dosages and purity of brands may vary.

Intra-articular Corticosteroids. Intra-articular corticosteroids have been shown to provide short-term relief of painful OA of the knee, but when results are analyzed more than 1 month after injection, the benefits are similar to those of placebo.²⁷ The rapid onset of action suggests that corticosteroids have a direct anti-inflammatory mechanism of action.³⁸ Intra-articular corticosteroids can be recommended when oral analgesics and NSAIDs are ineffective, for patients in whom NSAIDs are contraindicated, or when inflammation is a significant component of the patient's symptoms.²⁴ While some animal studies have suggested a link between weekly intra-articular corticosteroid injections and cartilage degeneration, historical data covering more than 330 000 injections found that this complication was less than 1%, well within the realm of chance.³⁹ Nonetheless, I would be reluctant to inject a knee with corticosteroids more frequently than every 3 or 4 months.

Intra-articular Viscosupplements. Hyaluronic acid, the major component of synovial fluid, becomes degraded in OA, leading to decreased viscosity and abnormal joint motion. Viscosupplementation is an attempt to enhance the viscous and elastic properties of abnormal synovial fluid, although an alternative mechanism is more likely.⁴⁰ Several industry-sponsored studies found that both hylan (Synvisc; Wyeth Pharmaceuticals) and sodium hyaluronate (Hyalgan; Sanofi-Synthelabo) may be as effective as oral NSAIDs and more effective than placebo injection, but the benefits may disappear as soon as 3 to 12 months after treatment and local adverse reactions have reportedly occurred in 8% of patients.^{41 - 43} Two studies were placebo-controlled, blinded, multicentered, and appropriately randomized with 102 and 495 patients, respectively^{41 ,43} ; the third was a retrospective study of 458 knees with follow-up of 2.5 years.⁴² Despite producing symptom relief, viscosupplementation has not been shown to change the progression or extent of disease or enhance cartilage repair.⁴⁴

Acupuncture. In the United States, an estimated 1 million patients use acupuncture annually.⁴⁵ Acupuncture treatment for OA of the knee has provided some intriguing data, but further study is necessary. Even if it is shown to potentially reduce pain, acupuncture does not alter the structure or course of arthritis. One meta-analysis of 7 mostly randomized trials of acupuncture in a total of 393 patients found that acupuncture can reduce pain, but enhancement in function was less predictable.⁴⁶ The series had anywhere from 14 to 103 patients. Another randomized controlled trial of 73 patients assessing the value of 8 weeks of acupuncture treatments found improvements in pain and function sustained for 12 weeks, but noted that the most significant benefits were realized in those patients with early arthritis.⁴⁷

Arthroscopy. Arthroscopic surgery is often performed for patients with acute mechanical symptoms, such as those caused by a meniscus tear or a focal chondral delamination lesion. When these lesions are superimposed on underlying arthritis, arthroscopic debridement may relieve the source of mechanical dysfunction, but at best only partially relieve the arthritic symptoms. Nonetheless, arthroscopic surgery has low morbidity and may play a role in the treatment of young patients with OA that has not responded to nonsurgical interventions, but who may not be candidates for total knee arthroplasty. Arthroscopic debridement 13 years ago provided Mr V with some, albeit temporary, relief of his pain.

The results of arthroscopic interventions for arthritis have been variable and unpredictable,^{48 - 49} with 50% to 70% receiving some degree of symptomatic relief, but most often deteriorating beyond 2 to 3 years.^{50 - 53} One intriguing prospective randomized study of 180 patients found no difference at 24-month follow-up in patients treated with either arthroscopic lavage, debridement, or a placebo in which there was a sham skin incision but no violation of the joint capsule.⁵² The authors did not identify the extent of relief of mechanical symptoms that were attributed to other problems such as meniscal tears, loose bodies, or chondral delamination lesions, which often cause severe and acute pain with mechanical symptoms superimposed on the underlying arthritis-related symptoms.

Periarticular Osteotomy. Proximal tibial or distal femoral osteotomy may be considered for patients with arthrosis limited to the medial or lateral compartments with varus or valgus malalignment of the limb, respectively.^{54 - 55} These procedures aim to reduce load distribution through the arthritic tibiofemoral compartment. The ideal patient for periarticular osteotomy is a young, active patient with modest malalignment, osteoarthrosis limited to one compartment (unlike Mr V), and intact ligamentous stability.^{54 - 55} Patients aged 60 years or older are generally considered to be more optimally treated with an arthroplasty. The pain relief achieved with osteotomy is incomplete and the results deteriorate with time, so that by 5 years, 37% may require total knee arthroplasty and by 10 years, 50%.^{56 - 58} In many patients, periarticular osteotomy fails to meet expectations, providing incomplete symptomatic resolution and leaving many active patients dissatisfied. In light of the inherent difficulties in converting periarticular osteotomy to total knee replacement, these procedures are best reserved for a small group of carefully selected patients.⁵⁸

Unicompartmental tibiofemoral arthroplasty (Figure 2), in which only the medial or lateral compartments are replaced, was traditionally reserved for sedentary patients older than 60 with no significant deformity.⁵⁹ Recently, however, attitude toward unicompartmental arthroplasty has shifted, partly related to modifications that allow the surgery to be performed through a small incision that affords a more rapid recovery, less pain, and shorter hospitalization than a standard approach.⁶⁰ While these studies were not randomized and controlled, short-incision unicompartmental arthroplasty has reduced the time frame to independence on stairs by as much as 72% compared with those after total knee replacement (vs 41% with a standard incision unicompartmental replacement). Additionally, the ability to perform a straight leg raise and flexion to 70° was reduced by two thirds compared with patients after total knee replacement vs only a third in those patients after standard incision unicompartmental arthroplasty. Additionally, compared with tibial osteotomy, unicompartmental arthroplasty provides more complete pain relief, fewer complications, improved muscle torque measurements, and a more normal gait.^{61 - 62} And compared with total knee arthroplasty, unicompartmental replacement has better patient satisfaction, range of motion, and ambulation capacity.^{63 - 64} Many surgeons now prefer this procedure to osteotomy as an intermediate option in younger, more active patients with isolated unicompartmental OA, prior to total knee arthroplasty. Mid- and long-term follow-up data have proven this option effective, with 85% to 90% implant survivorship at 10 years,^{60 ,65 - 66} even in patients younger than 60.⁶⁷ Mr V is not a candidate for this because of his multicompartment involvement. The survivorship of unicompartmental arthroplasty approaches but does not equal that of total knee arthroplasty. Failures tend to occur from wear of the polyethylene insert and progressive arthritis in other compartments of the knee.

After 4 decades, total knee arthroplasty remains effective for the treatment of degenerative joint disease of the knee (Figure 2B). It is the gold standard treatment for recalcitrant arthritis of the knee, against which all other techniques are ultimately compared. The potential of knee arthroplasty to relieve pain is obvious, but other favorable outcomes such as reducing energy expenditure, enhancing cardiovascular fitness, and improving well-being and quality-of-life scores have been documented.^{68 - 69} In one study assessing cardiovascular fitness after total knee arthroplasty, significant improvements were noted in maximum oxygen consumption (P = .008), maximum oxygen consumption corrected for body weight (P = .005), and percentage of predicted maximum uptake of oxygen (P = .005). Each of these demonstrates the increased physical activity capacity patients experience after total knee arthroplasty compared with a control group.⁶⁸ Quality-of-life measurements have been performed by several investigators and have shown significant improvement in patients' scores after total knee and hip arthroplasty compared with their preoperative status (P = .006).⁶⁹

Total knee arthroplasty is generally recommended for patients with advanced arthritis of the knee involving any of the 3 compartments of the joint, alone or in combination, when the arthritis is debilitating and has failed to respond to nonoperative interventions. It is also a useful option in the treatment of osteonecrosis. Total knee arthroplasty is not advised for patients with active or recent infection, in the setting of substantial extensor mechanism weakness, and arguably in neuropathic arthropathy or in extremely active young individuals.^{70 - 71}

Understanding patients' expectations is important to assess whether their objectives are realistic, especially for a patient like Mr V, who is an avid athlete. Knee replacement candidates usually hope for pain relief, improvement in performing daily activities, walking, kneeling ability, return to recreational activities, and an ability to climb and descend stairs comfortably.^{72 - 73} Physicians' objective outcome measurements of total knee arthroplasty may overestimate patients' perception of relief and success. Visual analog scales sometimes show persistent patient dissatisfaction despite good objective measurements, elaborating a disconnect between objective and subjective outcomes,⁷⁴ likely the result of unmet expectations.

Due to a combination of accelerated rehabilitation protocols and pressures from third-party payers, length of hospitalization has been reduced to 3 or 4 days for most patients undergoing unilateral total knee arthroplasty,⁷⁵ but patients often receive additional inpatient therapy before returning home. Crutches or a walker are used for the first 6 weeks and a cane for the next 6 weeks or until recovery of quadriceps strength. In addition to adequate strength, restoring motion is key to regaining function. Physical therapy should focus on restoring muscle strength, motion, and dynamic balance. A continuous passive motion machine used in the early postoperative period may accelerate recovery of motion.⁷⁶ Generally, mean range of motion after total knee replacement, 115°, is achieved by 3 months after surgery. While the majority of recovery occurs within the first 3 months, improvement in pain, motion, and function continue for up to 12 months after surgery.

Lifetime of Replacement. The quandary of total knee arthroplasty is that its long-term success has encouraged its use in younger more active patients and elderly recipients who are leading more active lifestyles, placing more stress on the implant. If one considers that implant wear is more a function of activity and repetitive cycling than age or obesity, per se,⁷⁷ then a balance between the desire of patients who are newly pain-free to perform unlimited activities and the prudent participation in activities is paramount. Most orthopedic surgeons recommend nonimpact or low-impact aerobic activities, such as swimming, walking, bowling, cycling, golfing, cross-county skiing, and doubles tennis. Running, downhill skiing, pivoting, and impact-type sports are generally discouraged.⁷⁸ Many return to the low-impact athletics listed above.^{79 - 80} Mr V could expect to resume cycling, but he should considerably reduce his mileage and gear resistance.

Refinements in surgical technique, improvement in implant design, metallurgy, quality of polyethylene, manufacturing, and sterilization methods of the polyethylene bearings have contributed to the remarkable performance of total knee arthroplasties, with survivorship greater than 90% at 10 to 15 years.^{81 - 85} Excellent mid-term results can be anticipated in several unique groups of patients, including both young and elderly patients,^{79 ,86 - 88} but in those with steroid-induced osteonecrosis⁸⁹ or after periarticular fracture about the knee^{90 - 92} the results are less predictable, with more complications, such as stiffness, malposition, premature loosening, pain, infection, and wound problems.

Potential Complications of Total Knee Arthroplasty. The risk of mortality within 30 days after total knee arthroplasty is less than 0.3%, and it is more likely in patients with preexisting cardiovascular and/or pulmonary disease, those older than 70 years, or patients with bilateral knee arthroplasties.^{93 - 94}

After total knee arthroplasty, the most common acute complication is venous thromboembolism. With contemporary prophylactic measures, the incidence of clinically significant thrombi has been reduced to 8% and fatal pulmonary embolism has been reduced to 0.05% to 0.2%.^{95 - 99} The risk of long-term sequelae is uncommon after deep vein thrombosis in total knee arthroplasty.¹⁰⁰ The risk of infection is less than 1.8% in the United States, and in most centers it is less than 1%.^{101 - 102} Perioperative antibiotics have been extremely effective for reducing infection, but several other intraoperative and postoperative strategies can further reduce the risk of infection.¹⁰³

Aseptic loosening and significant prosthetic wear occur infrequently in the first decade after surgery, but more commonly between 10 and 20 years. Characteristic prodromal signs and symptoms of failure that should raise suspicion of a failing total knee arthroplasty¹⁰⁴ include pain, swelling, progressive deformity, instability, stiffness, and clicking. Peroneal nerve injury has been reported in 0.3% to 2% of patients after total knee arthroplasty.^{96 ,105} Arterial complications have been reported to range from 0.03% to 0.3%, but the true incidence is difficult to establish accurately.^{96 ,106 - 107} Periprosthetic fractures of the distal femur, proximal tibia, and patella are most commonly associated with advanced osteoporosis, neurologic disorders, rheumatoid arthritis, and long-term corticosteroid therapy.⁹⁶

Mr V's condition poses a dilemma. He is extremely active but has considerable pain while performing his daily tasks, especially standing and walking on hard surfaces for prolonged periods. He is no longer able to jog because of advanced tricompartmental arthritis. Considering Mr V's young age and activity level, I would initially offer him an intra-articular corticosteroid injection and a trial of glucosamine and chondroitin sulfate, since these are the only medications he has not yet tried. If these interventions fail to relieve his pain, then total knee arthroplasty would be appropriate, but his expectations will need to be discussed at length. Mr V has displayed a tremendous level of athletic aptitude. Although I would certainly encourage nonimpact aerobic activities postoperatively, these will need to be done in moderation.

A PHYSICIAN: If this patient were to maintain his lifestyle and was able to cycle 200 miles a week but ultimately had failure of his knee earlier than would be anticipated, what's the likelihood that a repeat knee replacement would be successful?

DR LONNER: A lot depends on the integrity of the bone stock and the ligaments at the time of revision. If failure is recognized early, before a cascade of osteolysis and bone loss has occurred, successful revision can be anticipated. The results of revision knee replacement are rarely as satisfactory as they are after a primary knee replacement. If we look at mid-term data at 5 to 10 years, the global success rate of all revisions, all comers, is about 75% to 85%.¹⁰⁸ But this success is ultimately dependent on the quality of bone stock, integrity of the ligaments and soft tissues, skill and experience of the surgeon, and the type of implant used.

A PHYSICIAN: In terms of proprioception, I don't see physical therapists doing much balance rehabilitation when working with knee patients. Should we be more directive in our instructions?

DR LONNER: I think physicians should be descriptive when prescribing physical therapy. Specifically address range of motion, strengthening, and proprioceptive exercises because each element is important for functional recovery. In reality, though, after total knee replacement, many of the normal proprioceptive mechanisms within the joint are lost, so despite our directives, recovery of proprioception after total knee arthroplasty may be compromised.

A PHYSICIAN: In terms of long-term outcome after joint replacement, the major risk factors for failure are being healthy, being an athlete, and being male. What do you say to patients and how do you work on modification of behavior?

DR LONNER: Clearly there's a balance. We perform this procedure to allow patients to return to some semblance of a normal life. We want them to become more active once again. However, you have to counsel them that increased cyclic loading of the polyethylene can cause the replacement to fail. If a patient such as Mr V wants to go out and perform nonimpact aerobic activities or ride a bicycle you should encourage it, but I would not encourage cycling 200 miles a week. It is hard to set definite limits, but you do want to set some boundaries.