Patellofemoral Arthritis

The patella is a sesamoid bone, the largest in the body, embedded in the quadriceps and patella tendons. A subcutaneous bursa separates the patella from the overlying skin. The articular surface of the patella consists of four facets: inferior, middle, superior, and medial vertical.

During flexion, the patella moves within a groove in the femur, the trochlea. The patella is thought to travel in a J-shaped pattern, moving laterally with knee extension.[4]

The lateral aspect of the trochlea, the vastus medialis, and the medial patella-femoral ligament prevent excessive lateral translation. The anatomy of the distal femur and the vastus lateralis and lateral patella-femoral ligaments provides restraints against medial subluxation.

The patellar plexus provides the blood supply to the patella. The plexus is an anastomosis of the superior and inferior genicular arteries, which are branches of the popliteal artery,

Articular cartilage in the patella differs from that of other joints in several ways. The patellar cartilage is not necessarily congruent with the contours of the underlying subchondral bone. In 60% of patellae, the thickest area of articular cartilage is located lateral to the thickest area of underlying bone.

Some biomechanical studies have also indicated that this cartilage is less stiff and, thus, more compressible than that of other joints. One cadaveric comparison of femoral and patellar articular cartilage showed that patellar cartilage had a 66% higher permeability, a 30% lower compressive aggregate modulus, and was 23% thicker.[5] The proteoglycan content was 19% higher in the femoral cartilage than in the patellar cartilage. Whether these differences help prevent or promote arthritic changes in the patellofemoral joint is not yet clear.

Using magnetic resonance imaging, Herberhold and colleagues studied the relationship of static loading to cartilage deformation in fresh-frozen cadaveric knees.[6, 7] After 214 minutes of static loading with 150% body weight, the thickness of the articular cartilage was reduced by 44% in the patella and 30% in the femur. The changes in the thickness of articular cartilage of the patella were greatest at the lateral facet, the area of thickest cartilage.

Of note, 7% of the final deformation occurred during the first minute, and 25% occurred in the first 8 minutes. In the initial response to loading, the cartilage appeared to be stiffer than it was in response to prolonged loading. In normal loading, fluid in the cartilage is thought to support the applied load and to prevent cartilage deformation from occurring.

The main function of the patella is thought to be improving the mechanical advantage of the quadriceps extensor mechanism by increasing the lever arm of the muscle. The patella also acts to dissipate the forces generated in the patella tendon during knee flexion and extension.

The angle of force of the quadriceps muscle group (ie, the Q angle) is thought to be a factor in the development of knee injuries and arthritis. However, no findings conclusively support this assertion. The Q angle is measured as the angle between a line connecting the patella to the tibial tubercle and a second line between the anterior superior iliac spine and the center of the patella. A larger Q angle is thought to increase the lateral tracking of the patella mechanism.[8]

Huberti and colleagues concluded that, at normal Q angles, pressure is evenly distributed across the patella.[9] Increases in this angle, however, result either in a shifting of pressure to the lateral facet or a change in the distribution of force. Cadaveric studies demonstrate that with an increasing Q angle, the patella shifts more laterally and rotates medially as the knee is flexed. This change is thought to increase lateral contact at the patellofemoral joint and, possibly, to increase the incidence of patella subluxation and dislocation.

In closed-chain exercises with the foot planted on the floor (eg, squatting), contact forces increase with progressive degrees of knee flexion. However, in open-chain exercise in which the foot is off the floor (eg, hamstring curls), no corresponding increase in patella contact force occurs as the knee progresses through a range of motion.

The portion of the patella that is in contact with the trochlea also changes during range of motion.[10] With the knee extended, only the distal aspect of the patella contacts the trochlear groove. With progressive flexion, the contact shifts to the proximal aspect of the patella. At greater than 90° of flexion, the contact area is predominantly in the center of the patella, which corresponds to the thickest area of articular cartilage.

The medial facet of the patella articulates with the trochlea only during positions of full flexion. The force on the patella increases with knee flexion from 0-60°. However, no consensus exists among researchers regarding the relative amount of force generated with progressive flexion. The contact forces likely are related to the amount of force being generated by the quadriceps muscles during deep flexion exercises. Gait lab analysis has shown that walking with the foot plantar flexed, as occurs when high-heeled shoes are worn, increases the forces in the patellofemoral joint and in the medial compartment of the knee.

Patellofemoral arthritis can be a result of inflammatory conditions or mechanical abnormalities. Inflammatory conditions include rheumatoid arthritis; often, the entire knee joint is involved. Mechanical abnormalities can be a result of prior fractures, inherent malalignment, muscle imbalances, or chronic instability.

Chronic posterior cruciate ligament injury can lead to instability and pressure on the patellofemoral joint, causing arthritis and pain.

Patellar dislocation is a significant risk factor for patellofemoral arthritis. In one cohort study of 609 patients,  nearly half of patients had symptoms and radiographic changes consistent with arthritis at 25 years after lateral patellar dislocation.[11]

Association with anterior cruciate ligament reconstruction

Some orthopedists believe that reconstruction of the anterior cruciate ligament (ACL) with a patella tendon graft may lead to subsequent patellofemoral pain, loss of motion, and arthritis. However, whether these late complications are a result of the injury itself, the anatomic alignment that contributed to the injury, or the surgical reconstruction is not clear.

In an animal model, ACL transection caused significant changes in joint pressure and the articular cartilage after only 4 months.[12] The thickness of the articular cartilage significantly increased in the patella, the medial and lateral femoral condyles, and in the patellar groove, with total increases in the range of 42-100%. Although the joint contact area was increased, overall peak contact pressures were reduced. The thickening of the cartilage may provide a protective modification of joint pressure, or it may represent early arthritic changes.

In a retrospective clinical study, Jarvela et al found an association between prior ACL reconstructions with bone-patella-bone allografts and the subsequent development of patellofemoral arthritis.[13] Although evidence of arthritis along the lateral and medial tibial femoral joints was found in 15% and 18% of these patients, respectively, 47% had radiographic evidence of patellofemoral arthritis.

At follow-up at a mean of 7 years, patellofemoral arthritis was mild in 34%, moderate in 12%, and severe in 1%. Shortening of the patella tendon after the index procedure was significantly associated with the development of arthritis. However, the location of the bone tunnel did not correlate with the development of arthritis.

In a cross-sectional study of 70 patients who underwent hamstring tendon ACL reconstruction 5-10 years previously, Culvenor and colleagues found that patellofemoral arthritis was common and was associated with greater knee-related symptoms and impaired functional performance. Radiographic evidence of arthritis was observed in the patellofemoral joint in 47% of patients and in the tibiofemoral joint in 31%. Measures of pain, other symptoms, and quality of life were associated with the severity of patellofemoral arthritis. Medial meniscal and patellofemoral chondral lesions at the time of surgery were associated with the development of tibiofemoral and patellofemoral arthritis. A longer surgery delay was associated with the development of patellofemoral arthritis.[14]

Patellofemoral joint osteoarthritis and arthritis in other knee compartments

Osteoarthritis of the patellofemoral joint should be considered as an entity separate from disease in the medial and lateral tibiofemoral compartments of the knee. Not all patients with patellofemoral arthritis have osteoarthritis in the other compartments; arthritis may develop at different times and with different etiologies in the different compartments of the knee.

Some studies have indicated that risk factors for the development of patellofemoral arthritis include increasing body weight, high-intensity running or weight lifting, prior knee injury, and prior patellar dislocation or subluxation.

Kujula and colleagues conducted a longitudinal study to address the mechanical factors leading to arthritis of the patellofemoral joint[15] and found that of those with radiographic evidence of patellofemoral arthritis, 70% had changes to the lateral side of the patella. In this subset of patients, the likelihood of a varus knee alignment was significantly increased compared with those with medial patellofemoral arthritis, who were more likely to have a valgus knee. A total of 292 patients with osteoarthritic knees were evaluated radiographically.

In one third of the patients with patellofemoral arthritis, no radiographic evidence of osteoarthritis was present in other compartments of the knee. This isolated patellofemoral arthritis was seen more often in a valgus alignment than was isolated tibiofemoral arthritis.

In another study, computed tomography of 40 knees in flexion revealed that in patients with patellofemoral arthritis, the tibial tubercle had a significantly more lateral position than it does in normal knees.[16] However, whether this finding represents a cause or a result of arthritic damage to the joint was not determined.

Chondromalacia

In 1962, Outerbridge observed that half of his patients had evidence of irregularities in the articular cartilage of the patella at the time of meniscectomy.

Patellofemoral arthritis

Approximately 5% of patients with osteoarthritis (OA) of the knee have symptomatic patellofemoral arthritis in the absence of tibiofemoral arthritis. The etiology of the arthritis is divided equally among patellar dislocation, fracture, and primary OA.

A systematic review and meta-analysis of 85 studies that reported the prevalence of patellofemoral OA and structural damage found that approximately one half of people with knee pain or radiographic OA have patellofemoral involvement. Radiographic evidence of patellofemoral OA was present in 43% of those with knee pain or symptomatic knee OA and in 57% of those with radiographic and symptomatic knee OA.[17]

Patients with patellofemoral arthritis typically present with anterior knee pain. The patient should be asked the following questions:

• How long has the pain been present?

• What makes it worse?

• Is the condition aggravated by prolonged squatting, stair climbing, or other activities?

• Is the pain dull and achy or is it sharp?

• Have you sought treatment for this condition in the past?

• Why are you seeking treatment now?

• Has there been a recent change in your activity?

• What type of work do you do?

• What other types of activities do you participate in (eg, gardening, kneeing at church, yoga, cycling)

• Have you noticed swelling in the knee?

• Have you had prior knee surgeries?

Pain from arthritis and malalignment is typically variable, becoming worse with activity. Constant pain that does not vary with activity suggests a referred or nonmechanical origin.

Isolated patellofemoral arthritis may cause anterior knee pain that worsens with stair climbing or when rising from a seated position and is not present with other activities, such as walking or running on level surfaces.

Patella instability is associated with intermittent sharp pain at the kneecap. A feeling of "giving way" may be related to muscle weakness or to instability. Recurrent patellar subluxation or dislocation may cause an acute osteochondral fracture or chronic cartilage damage as a result of repeated microtrauma.

The physical examination comprises observation, measurement, and palpation.

Observation

Astute observations of gait and of lower-extremity alignment are an essential component of the physical examination. Leg-length discrepancies, Q angle, and torsional deformities of the femur, tibia, and foot should be noted. Flexion contractures of the limb should be noted.

Gait, such as waddling gait, should be observed carefully, with the patient not wearing shoes. Excessive pronation of the feet, patella tracking, and rotation of the lower limb should be observed. Muscle tone and atrophy of the quadriceps and hamstrings should be assessed. Patella tracking with passive flexion and extension and with very careful semisquatting should be determined.

A study by Crossley et al found that, compared with normal controls, people with patellofemoral joint osteoarthritis demonstrated the following differences while walking[18] :

• Greater anterior pelvic tilt throughout the stance phase,
• Greater lateral pelvic tilt (ie, pelvis lower on the contralateral side)
• Greater hip adduction
• Lower hip extension during the late stance phase

Physical findings must be evaluated in the context of patient complaints. A study of 210 adults with asymptomatic knees revealed abnormal radiographic or physical examination findings in 95% of women and 79% of men.[19] Patellar crepitus, a hypermobile patella, and lateral position of the patella on the axial radiographs were common in this group of patients. Unfortunately, no long-term follow-up was performed to determine whether subjective complaints developed; however, these findings emphasize the point that a diagnosis cannot be based on physical findings alone.

Measurement

Measurements should be taken of femoral anteversion, knee valgus, tibial pronation, lower limb length, and Q angle.

Limb length is measured from anterior iliac spine to the medial malleolus. Length discrepancies should be noted.

The Q angle is the angle between an imaginary line extending from the anterior superior iliac spine to the patella and a line from the patella through the shaft of the tibia. Normal values are less than 20°. Women typically have larger Q angles than men because of their wider hips. While an association is thought to exist between varus knee alignment and the development of osteoarthritis, no study has ever definitively linked Q angles to knee pathology.

Palpation

A thorough examination of both the affected and nonaffected knees should be performed. The presence of crepitus is nonspecific. A standard knee examination should be performed. Passive and active range of motion should be recorded. Strength and tightness of hamstring and quadriceps muscle groups should be determined.

Assess for the presence of ligamentous laxity, instability, and patella maltracking and attempt to elucidate the source of pain. The goal of palpating the structures of the anterior knee is to determine whether the patient's complaints are related to arthritic changes or to the underlying soft tissues.[20] Attention should be focused on the lateral retinaculum, the quadriceps and patellar tendons, and the quadriceps muscle.

The patella is compressed as the patient flexes the knee. Pain often is elicited by this maneuver if arthritis is present. Resisted knee extension also may reproduce the patient's symptoms in arthritic conditions.

Attempt to laterally displace the patella with the knee in extension. Patients with instability contract their quadriceps muscles or complain of pain because of the feeling of subluxation

Knee stability to varus and valgus stress should be assessed. Stability of the ACL can be determined by anterior drawer and Lachman tests. The stability of the patella to medial and lateral stress should be determined, as should the lateral patella tilt.

The patella normally enters the trochlea from a lateral position and becomes centralized with increasing knee flexion, traveling in a J pattern. Abnormalities observed include excessive lateral tracking increasing the angle at which the patella enters the trochlea.

Perform a hip and spine examination. Additionally, look for a prominent superolateral patellalike bipartite patella or a high-riding patella.

Patellar tendinitis, which is swelling at the inferior pole of the patella tendon, is also a possibility and must be evaluated.

Radiography is the principal study used in the diagnosis of patellofemoral arthritis. Computed tomography can also be used to determine patellar tilt and articular damage. Arthroscopy can be used for both diagnosis and therapy.

No laboratory tests are usually required for the initial workup for knee pain. However, if clinical suspicion arises, a workup for systemic causes of polyarticular arthritis may be appropriate. This workup includes but is not limited to an evaluation for Lyme disease, rheumatoid arthritis, psoriatic arthritis, and gout.

Radiographic views of the knee include standing anteroposterior (AP), lateral, and axial (Merchant view, 45° flexion).[21] These radiographic views should be examined closely for narrowing of the joint space, osteophytes, articular degeneration, and patella alignment. Congruence of the patella and patella tilt can be assessed on the axial view. Specific measurements, which can be obtained from the radiographs, include the sulcus angle, the congruence angles, and the Insall-Salvati ratio (described below).

A sulcus angle is the angle between the medial and lateral condyles and the sulcus. It normally averages 138°. (See the image below.)

Patellofemoral arthritis. The sulcus angle is the angle formed by the condyles and the sulcus. This is a measurement of trochlear depth. A finding greater than 138° indicates the presence of patellofemoral dysplasia.

A congruence angle is the angle between a line bisecting the sulcus angle and the articular ridge. It normally averages 6° and is a measurement of subluxation.

The Insall-Salvati ratio is the ratio of the length of the patella ligament to patella length as measured on the lateral view. The ratio should be 1:1; a ratio of more than 1:2 indicates the presence of patella alta. (See the image below.)

Patellofemoral arthritis. The Insall-Salvati ratio is used to assess the position of the patella and determine if the patella alta or patella baja is present. The ratio of the height of the patella to the length of the patella tendon should be 1:1.

The patellofemoral index is a distance ratio of the articular ridge and medial condyle to the articular surface and the lateral condyle. (See the image below.)

Patellofemoral arthritis. The patellofemoral index is the ratio of M, the closest distance between the articular ridge and the medial condyle, and L, the closest distance between the lateral facet and condyle.

With radiography, the presence of osteophytes at the lateral femoral trochlea combined with narrowing of the joint has 90% sensitivity in diagnosing arthritic changes, as compared with MRI.[22]

Arthroscopy is an excellent tool for diagnosing patellofemoral arthritis.[22] The surface of the patellofemoral and tibiofemoral joints can be evaluated under direct visualization to assess the size, severity, and location of articular degeneration. The motion of the patella with flexion and extension can be observed to determine whether subluxation is present. Arthroscopy can also be used for therapy.

Normal articular cartilage is smooth and glistening. Degenerative changes observed in the articular cartilage of the patella are similar to those observed in other joints. In one of the earliest descriptions of the pathology of the patellofemoral joint, Outerbridge described dulling of the color and softening of the cartilage as the earliest changes in chondromalacia patellae.[3] With progressive degeneration, osteoarthritis develops, and fissures and erosion of the cartilage occur. These changes are typically more pronounced on the medial patellar facet than elsewhere, but eventually extend throughout the patella.

Nonoperative treatment involves nonsteroidal anti-inflammatory drugs (NSAIDs), preferably topical; activity modification; and muscle-strengthening activities. Indications for surgical management of patellofemoral arthritis include the following:

• Pain
• Loss of functional ability
• Arthritis that correlates with the symptoms
• Symptoms that do not respond to physical therapy

Surgical treatment can begin with arthroscopic assessment of the patellofemoral articular cartilage. If symptoms persist, tibial tubercle osteotomy can be considered for patients younger than 50 years of age and older active patients. For patients with severe, refractory patellofemeroal joint osteoarthritis, patellofemoral arthroplasty should be considered.[23]

Go to Rheumatoid Arthritis for more complete information on this topic.

Strengthening of the quadriceps muscle of the quadriceps muscle group can be beneficial, especially in patients with maltracking and weakness of these muscles.[24] Activities should be modified so that prolonged flexion (as with squatting) and stair climbing are avoided. A knee sleeve may also relieve symptoms. Weight loss is beneficial in overweight or obese patients.

Some have advocated taping of the patella to push it into a more medial position, especially in young patients undergoing rehabilitation for anterior knee pain.[25] Interestingly, a study that used computed tomographic scanning to evaluate patellar position found that taping had absolutely no effect on either patellar displacement or patellar tilt.[26]

A prospective, randomized, double-blind study of nonoperative treatment in younger patients (aged 20-55 y) with patellofemoral pain (not arthritis) revealed that symptoms resolved in 67% of patients within 6 months of initiating physical therapy, and 80% graded their knee as excellent after 7 years. However, clinical findings in these patients worsened with time. Significantly more patients had positive apprehension and compression test results and had crepitus at 7-year follow-up, and 5% of patients developed arthritis.[27]

A single-center, randomized, controlled trial of arthroscopic surgery in patients with moderate-to-severe knee osteoarthritis found that surgical lavage and arthroscopic débridement together with optimized physical and medical therapy provides no additional benefit compared with optimized physical and medical therapy alone. After 2 years, the total Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) score for the surgery group was 874 +/-624, as compared with 897+/-583 for the control group (P=0.22).[28]

Although conservative management may help in resolving symptoms, no evidence indicates that it prevents the progression of arthritic changes in the joint.

Before proceeding with surgery, the surgeon and patient must be in agreement regarding the goal of the procedure. Goals may include correction of malalignment, balancing of soft tissues, and assessment and treatment of articular damage. The major goal is to reduce pain and swelling and improve function.

Medical conditions such as hypertension, coronary artery disease, and diabetes should be managed and stabilized as much as possible prior to surgery. Medical clearance prior to surgery is recommended for patients with multiple medical issues.

The patient must understand the procedure. Additionally, patients must be willing and able to comply with postoperative rehabilitation regimens.

Surgical treatment of patellofemoral arthritis is elective. Patients should be in relatively good health prior to the operation. Contraindications to surgery include hemodynamic instability, current respiratory infection, recent myocardial infarction, and compromised skin integrity in the surgical field.

Arthroscopic debridement has been used to smooth and clean up fibrillated cartilage on the joint surface. No evidence indicates that this provides long-term relief of pain, however. Articular cartilage does not regenerate; instead, it is replaced by fibrocartilage. Although not as mechanically protective as articular cartilage, fibrocartilage may induce less of an inflammatory response than damaged articular cartilage.

Kirkley et al found that "arthroscopic surgery for osteoarthritis of the knee provides no additional benefit to optimized physical and medical therapy."[29] In an accompanying editorial, Marx stated, "However, osteoarthritis is not a contraindication to arthroscopic surgery, and arthroscopic surgery remains appropriate in patients with arthritis in specific situations in which osteoarthritis is not believed to be the primary cause of pain."[30] Also see the Medscape article "Arthroscopic Surgery May Not Be Helpful for Knee Osteoarthritis."[31]

Controversy exists about the benefit achieved from any kind of surgery and, specifically, arthroscopic surgery on a patient who has painful swelling from patellofemoral arthritis and in whom conservative treatments (eg, knee brace, physical therapy, anti-inflammatory medications) have failed.

In selected patients who have loose bodies and synovitis and swelling and those who are not willing to have total knee replacements, arthroscopic surgical debridement can be beneficial in the hands of an experienced surgeon; however, patients must be aware of the unpredictable benefit early after surgery or long term.

Providing clear information on this is the first step before any decision is made regarding total knee surgery, especially for patients who are not involved in any type of significant recreational activities or have medical problems that may prevent full completion of the total knee replacement surgery process.

For patients with anterior knee pain directly associated with poor patellar positioning, surgical procedures that attempt to correct the malalignment may be beneficial. The goal of these procedures is to decrease pain by changing the distribution of force across the patella.

A lateral retinacular release is performed to change the tilt of the patella, decreasing pressure on the lateral patellar facet by releasing the lateral constraints of the patella. These constraints include the iliotibial band, vastus lateralis, and patellofemoral and patellotibial ligaments. This procedure is often performed with patellofemoral and total knee arthroscopy when lateral patellar malalignment is noted.

In a study of 85 patients scheduled to undergo total knee arthroplasty caused by varus OA, researchers observed that the femorotibial angles in patients with stage II-IV patellofemoral OA were significantly larger than those in patients with stage I patellofemoral OA; the patellar tilt in patients with stage II-IV patellofemoral OA and the tibial tuberosity-trochlear groove (TT-TG) distance in patients with stage IV patellofemoral OA were also significantly larger than among patients with stage I patellofemoral OA. The TT-TG distance was strongly correlated with patellar tilt (R(2) = 0.41, P < 0.001). The reserachers concluded that prevention of patellofemoral OA may be facilitated with procedures such as high tibial osteotomy and total knee arthroplasty to correct knee malalignment.[32]

Medialization of the tibial tuberosity has been advocated to improve patella tilt and subluxation by decreasing the valgus moment acting on the patella. Anteriorization of the tibial tubercle is used in more severe cases and increases the power of the quadriceps by increasing the length of the lever arm. Patellofemoral joint reactive forces are decreased with this procedure.

The existence of multiple procedures to correct malalignment suggests that no single procedure consistently results in excellent outcomes. These procedures may be beneficial for young patients with intractable knee pain unresponsive to medical therapy or in some patients with early arthritic changes. For patients with severe patellofemoral arthritis, a more extensive surgical procedure is required to address the articular damage.

The ideal solution to articular cartilage damage is cartilage regeneration. Unlike bone, damaged articular cartilage lacks the ability to repair itself. Techniques such as subchondral drilling, microfracturing, and abrasion arthroplasty have been used to stimulate healing of the articular cartilage in the knee joint. Some have theorized that undifferentiated progenitor cells in the marrow are stimulated to form articular cartilage.

Histologic studies, however, have demonstrated that healing of the defect occurs primarily with fibrocartilage, which has inferior mechanical properties compared with those of hyaline cartilage. The limitations of these techniques spurred research into replacing the defect with chondrocytes or osteochondral autografts or allografts to restore the hyaline cartilage of the articular surface.

Autologous cartilage transplantation and osteochondral autografting

Briefly, the procedure of autologous chondrocyte transplantation involves harvesting chondrocytes from a relatively non–weight-bearing portion of the medial tibia condyle. The cells are cultured for 11-21 days, at which time they are implanted into the defect. The defect is covered with a periosteal patch from the proximal medial tibia to keep the cells in place. Patients must be non–weight-bearing for approximately 8 weeks to avoid damaging the transplanted cells.

Brittberg et al published their intermediate (39-mo) follow-up on 23 patients with full-thickness cartilage defects in the knee that were treated with autologous cartilage transplantation[33] and determined that of 16 patients with femoral condylar defects, 14 (87%) had good-to-excellent clinical results. Transplants to the patella region had less favorable results, with only 2 (29%) of 7 patients having good-to-excellent results. Hyaline cartilage was found on biopsy in 11 of 16 femoral transplants compared with 1 of 7 patellar transplants.

Peterson et al reported that 91% of 58 patients had good-to-excellent results, based on clinical examination and results self-assessment questionnaires.[34] In this series, however, all of the patients had femoral condylar defects and none was treated for osteochondral defects of the patella.

The limitations of this type of autologous cartilage transfer include the requirement of two surgical procedures, the risk of infection being introduced when the cells are grown in vitro, and the facilities and expertise needed for the successful in vitro growth of chondrocytes. A similar method, autologous osteochondral mosaicplasty is performed to achieve the same results without these limitations.

In this technique, multiple, small (2.7- to 8.5-mm diameter) plugs of hyaline cartilage and underlying subchondral bone are taken from non–weightbearing surfaces. They are then implanted into the osteochondral defect. The donor site fills with fibrocartilage, but, theoretically, the hyaline cartilage of the transplant should be preserved.

Patients are typically advised not to bear weight for 2 weeks following the procedure, and then partial weight bearing is allowed for 2 weeks. Hangody et al reported results in 831 consecutive patients who underwent this procedure during a 10-year study period, with complications including deep infection (n = 4) and painful hemarthroses (n = 36). Clinical evaluation with standard knee scores indicated good-to-excellent results following mosaicplasty in 92% of femoral (n = 597) procedures, 94% of talar procedures (n = 76), 87% of tibial procedures (n = 25), and 79% of patellar procedures (n = 118). Patients younger than 35 years had better results than those in older patients.[35]

According to Bentley et al, in a prospective randomized comparison of autologous cartilage transplantation and mosaicplasty,[36] at 19-month follow-up, 88% of patients (n = 58) with cartilage transplants and 69% of patients (n = 42) with mosaicplasty had good-to-excellent outcomes.

Both autologous cartilage transfer and mosaicplasty may hold promise for the treatment of patellofemoral osteochondral lesions. However, results for osteochondral lesions of the patellofemoral joint with these procedures appears to be less favorable than for lesions in the femoral condyles, and the procedures require careful patient selection.

These procedures are best suited for young, active patients who have focal degenerative lesions and in whom moderate-to-severe osteoarthritis has not yet developed. Additional surgical procedures to correct misalignment may also be necessary. Patients must be highly motivated and willing to cooperate with prolonged non–weightbearing periods.

Osteochondral allografting

Osteochondral plugs or wafers can also be obtained from allografts, which limit damage to the articular surface and increase the size of defect that can be repaired. Fresh allografts of articular cartilage and subchondral bone are implanted into the defect. The allograft bone is reabsorbed and remodeled, but the transplanted chondrocytes survive.[37, 38]

Chu et al reported that in 55 consecutive patients,[39] about 84% with unipolar transplants, but only 50% with bipolar transplants, had good-to-excellent results. They also reported that 11 (73%) of 15 patients had good-to-excellent clinical results at 10-year follow-up.

In a long-term follow-up study of osteochondral grafting for isolated posttraumatic tibial plateau defects, by Shasha et al,[40] Kaplan-Meier analysis revealed a survival of 95% at 5 years, 80% at 10 years, and 65% at 15 years, with an endpoint of knee arthroplasty, revision of graft, or a Hospital for Special Surgery knee score of less than 70.

A case series by Gracitelli et al of fresh osteochondral allograft transplantation for isolated patellar cartilage injury in 28 knees concluded that transplantation was successful as a salvage treatment procedure. Eight of the 28 knees (28.6%) were considered treatment failures. Graft survival was 78.1% at 5 and 10 years and 55.8% at 15 years.[41]

Although this technique does not violate the native hyaline cartilage of the knee, fresh allografts must be used, as freezing or other processing destroys the viability of the chondrocytes. This limitation has limited enthusiasm for this procedure because, despite precautions, the risk of bacterial, viral, or fungal disease transmission remained increased with this type of transplant.

Patellectomy is a surgical option after severe patella fractures. Surgeons have also used this procedure as a treatment for arthritis of the patella. Problems associated with patellectomy include loss of normal knee power and function, quadriceps weakness, and failure to resolve anterior knee pain. Cybex measurements after patellectomy demonstrate significant decreases in the peak torque of the quadriceps muscles, with preservation of normal hamstring function.[42] Patellectomy in a rabbit model has been shown to result in arthritis of the tibiofemoral joint.[43]

Clinically, a retrospective review of 81 patients who underwent patellectomy for osteoarthritis during a 20-year period found that only 53% achieved a good result.[44] With time, there was a statistically significant increase in radiologic changes at the tibiofemoral joint consistent with the development of arthritis. A case report of arthroscopic findings in 16 knees with a prior patellectomy showed severe medial compartment and trochlear articular damage in patients aged 22-64 years with a mean follow-up of 15.4 years.[45] With progressive degeneration, these patients may later require a total knee arthroplasty (TKA) for arthritis of the knee.

To the authors' knowledge, no large studies reveal whether patellectomy restricts future options for TKA. Joshi et al reported a complication rate of 36% in 19 patients with prior patellectomy compared with 0% in a matched cohort.[46]

A retrospective matched control study at the Hospital for Special Surgery examined the results of TKA in patients who either did or did not undergo previous patellectomy.[47] Only 6 (30%) of 22 patients underwent patellectomy because of osteoarthritis. At 5-year follow-up, standardized knee scores did not differ between the 2 groups of patients. However, patellectomy results did differ depending on the prosthesis used.

Patients with a posterior cruciate ligament (PCL)–sacrificing prosthesis had statistically higher postoperative knee scores than those with a PCL-retaining prosthesis. The subgroup of patients with PCL conservation had more pain, decreased range of motion, and more difficulty ascending stairs. The authors believe that the posterior stabilized implant increases the quadriceps lever arm. Considerable debate exists, however, regarding which implant provides better results after patellectomy, as other earlier groups have reported contradictory findings.

In 1955, McKeever performed the first replacement of the patella as an alternative to patellectomy for isolated patellofemoral arthritis. Since that time, the patellar prosthesis has been reengineered multiple times. The most commonly used implants are cemented polyethylene implants. The procedure can consist of a patella-resurfacing procedure alone or may include femoral resurfacing and the replacement of the femoral groove with a metal implant. Isolated patellar resurfacing, however, has not been shown to be more effective than total knee replacement (TKR) in the older patient population.[48, 49]

One significant advantage of arthroplasty over patellectomy is the preservation of the quadriceps function. Disadvantages of this procedure are that it is a technically demanding procedure and not commonly performed and that a second surgery may be needed for TKR if the patient also has tibiofemoral osteoarthritis. Some concern also exists about the effect of particulate metal debris collecting in the joint from the metal-backed component, which may be associated with an increased risk of infection.[50]

Krajca-Radcliffe and Coker reported a retrospective study of 16 total patellofemoral arthroplasties in 13 patients conducted between 1975 and 1992 with a minimum 2-year follow-up and average patient age of 64 years.[51] About 88% had good or excellent results, 7 of 13 were pain-free with all activities, and no patients had pain with routine activity. Quadriceps strength was 75% and 100% of the contralateral side. One patient required a revision procedure.

In a larger series of 72 patellofemoral arthroplasties in 65 patients with an average 4-year follow-up, 85% had good-to-excellent results. However, more than half the patients in this series had concurrent unicompartmental knee replacements. Furthermore, the authors reported complications in 14 patients (18%), 5 of whom required a repeat operation.[52]

Another review of 66 isolated patellofemoral arthroplasty procedures performed over a 10-year period with a 5.5-year average follow-up demonstrated that 16% required revision to a TKA.[53]

Problems related to patellar maltracking plagued earlier-generation patellofemoral arthroplasties. These problems occur less often with contemporary components.[54]

A systematic review and meta-analysis comparing modern patellofemoral arthroplasty with TKA for treatment of isolated patellofemoral osteoarthritis concluded that, in appropriately selected patients, patellofemoral arthroplasty seems to be a viable alternative to TKA. Patients undergoing patellofemoral arthroplasty had less operative blood loss, shorter postoperative hospital stays, and similar patient-reported outcome measures.[55]

TKA is used to treat patients with severe osteoarthritis of the knee that has not improved with medical management and that has become detrimental to a patient's quality of life. It is usually performed in patients older than 60 years, but occasionally, younger patients with significant arthritic changes are considered candidates for the procedure.

At the Hospital for Special Surgery in New York, Laskin and van Steijn compared the results of TKA in 53 patients with severe patellofemoral but minimal tibiofemoral arthritis to sex- and age-matched patients undergoing the procedure for tricompartmental knee arthritis.[56] All patients had a cemented tricompartmental PCL-sparing prosthesis. Preoperatively, the group with isolated patellofemoral arthritis had statistically significant increases in pain score, and fewer were able to rise from a chair or climb stairs without assistance.

In this study, lateral release was required in 30% of patients with patellofemoral arthritis versus 8% in the other group. Postoperatively, 82% of the patients were able to climb stairs, and only 3 complained of anterior knee pain. Average range of motion was 120°. These results were comparable to those in the other group. This study indicates that TKA may be the procedure of choice in older patients with severe patellofemoral arthritis.

Bicompartmental knee arthroplasty versus TKA

In North America, knee osteoarthritis is 5-10 times more likely to involve the medial compartment than the lateral compartment. Although TKA is the standard of treatment for osteoarthritis of the medial and patellofemoral compartments that has failed conservative treatment, bicompartmental knee arthroplasty (BKA) has drawn increasing attention because it preserves proprioception and native knee kinematics.[57]

A systematic review and meta-analysis of six studies that included 274 patients and 277 knees concluded that the use of BKA (with a modular rather than a monolithic implant) for medial compartment and patellofemoral osteoarthritis is comparable with TKA in terms of short-term function, complication rate, and revision rate. Intraoperative blood loss is lower with BKA but operation length is longer, due to its greater technical demands.[57]

Patellar resurfacing versus nonresurfacing in TKA

The need to perform a patella resurfacing procedure at the time of TKA is still being debated. Some surgeons advocate resurfacing in all patients, while others restrict resurfacing to patients with known patellofemoral arthritis.

Patellofemoral complications after TKA with patella resurfacing are associated with obesity, metal-backed patellar implants, and noncemented patellar components.[58] These complications can include pain, mechanical fracture, loosening, fracture, and maltracking.

Retaining the native patella, however, can result in persistent symptoms of anterior knee pain that can compromise the functional ability of the patient. Furthermore, patients with known osteoarthritis in the tibiofemoral knee compartment are at increased risk for arthritis of the patellofemoral joint.

Some surgeons have suggested that patella resurfacing does not have to be performed if the patellar articular cartilage is normal and if no malalignment is present. The results of several small retrospective studies support the practice of selective resurfacing. In a retrospective study of 185 patients with TKR, pain, functional outcome scores, range of motion, or complications did not differ between those undergoing resurfacing and those not (n = 140).[59]

In a 10-year retrospective follow-up study in 32 knees with a PCL-retaining Press-Fit condylar TKA and no resurfacing, only 1 patient required later revision for patella resurfacing.[60]

Several investigators have asserted that no difference exists between resurfacing and retaining the patella in terms of stair-climbing ability or functional outcome.[61, 62] Furthermore, if the patella is resurfaced and if a complication requiring revision is present, the patella often cannot be resurfaced again because of bone loss. For this reason, some surgeons do not recommend resurfacing in young patients, even when some patellar chondromalacia is present. If the patella is left unresurfaced, the knee system should have an anatomic rather than dome-shaped trochlear groove in the femoral component.

Other orthopedic surgeons believe that a resurfacing procedure should be performed in all patients at the time of a knee arthroplasty. Biomechanically, most prosthetic knees are designed for use with a resurfaced patella.

A study examining the relationship of the patella to different femoral components revealed that retaining the patella results in changes in contact pressures and in the tracking motion in the patellofemoral joint.[63] With flexion of 60° or more, the contact areas shifted from a transverse band across the patella to isolated areas of high pressure on both the medial and the lateral sides of the patella.

All of the prostheses had a significantly greater percentage of patellar contact area subjected to contact pressures above 5 MPa compared with the normal knee with flexion greater than 60°. The patella tracked 3-5 mm more laterally once a prosthesis was in place. This finding may help to explain why some patients without patellofemoral arthritis who have a retained patella may have more difficulty climbing stairs postoperatively than those who undergo a resurfacing procedure.

In a prospective series of 40 patients, those undergoing patellar resurfacing had clinically improved function at a 2-year follow-up in terms of stair climbing and higher functional k scores than those of patients not receiving this procedure.[64] One group in Japan that routinely performed TKA without resurfacing the patella concluded that patellar resurfacing should not be performed routinely but, rather, should be considered for patients with rheumatoid arthritis, who are vulnerable to postoperative thinning of the patella and peripatellar pain.[65]

Boyd and colleagues[14] examined the results of using an unconstrained, condylar, PCL-preserving prosthesis with and without resurfacing in 891 knees and found that the rate of complications was higher in those without resurfacing than in those with resurfacing (4% vs 12%). Patella resurfacing was performed only for specific conditions: loss of cartilage, surface irregularities, and tracking abnormalities. Of patients without resurfacing, 10% required revision for patella resurfacing because of chronic pain. Those with inflammatory arthritis were significantly more likely to require revision for patellar pain than those with osteoarthritis (13% vs 6%).

To the authors' knowledge, no large prospective studies currently provide clarity in this debate. Patients with rheumatoid arthritis seem to do better with a resurfacing procedure. To decrease the incidence of a subsequent revision surgery, most patients would most likely benefit from patella resurfacing at the time of TKA. Functional outcomes, however, seem to be similar when selective nonresurfacing is performed.

In a retrospective study of 39 knees in 28 middle-aged to elderly patients with predominant lateral patellofemoral osteoarthritis, partial lateral facetectomy, a minimally invasive procedure, improved initial anterior pain relief, as measured with the Knee Society Score, in 84% of knees. However, 30% of the knees eventually required total knee replacement.[66]

The type of follow-up care required depends on the type of surgical procedure performed. Postoperative physical therapy focused on quadriceps strengthening is helpful to many patients. Those undergoing TKA or patellofemoral arthroplasty may start range-of-motion exercises immediately after surgery, and, depending on their preoperative ambulatory status, patients may require a short stay in a rehabilitation hospital while they receive physical therapy.

Procedures involving anteriorization of the tibial tubercle require a period of relative immobilization after surgery and before physical therapy. Cartilage transplantation and osteochondral grafts require a period of strict non–weight bearing for several weeks to allow for the incorporation of the graft. The patient's understanding of and compliance with postoperative treatment protocols are important in the success or failure of all of these surgical treatments.

Complications of patellectomy include quadriceps weakness, loss of normal knee function, and failure of the procedure to resolve the patient's pain. Some studies indicate that patellectomy may result in an increased incidence of knee arthritis. If osteoarthritis develops later, the patient may require a second surgery for TKA, and they may have a greater-than-expected complication rate after this procedure.

Complications of patellofemoral arthroplasty include patellar fracture, infection, and adverse reactions to metal debris in the joint. Revision to a total knee occurs in an average of 16% of patients.

Complications of TKA include the infection, loosening, patella fracture, and periprosthetic fracture.

All surgical procedures have risks associated with anesthesia. The extent of the risk is related to the general medical condition of the patient. Knee surgeries are associated with a risk of deep venous thrombosis, which is reduced with the use of appropriate prophylaxis. Pulmonary embolism from propagation of thrombus after total knee reconstruction (TKR) occurs in less than 1% of patients.

Surgical therapy is beneficial for patients with intractable pain and radiographic evidence of arthritis. Patellectomy, patellofemoral arthroplasty, and TKA reduce pain and increase function. Patellectomy, however, is associated with significant changes in joint biomechanics, and one study demonstrated that only 53% of patients achieved a good result at long-term follow-up.

Nonoperative measures for patellofemoral arthritis are primarily aimed at relieving pain and swelling. Nonsteroidal anti-inflammatory drugs (NSAIDs) are commonly used to control inflammation.

Class Summary

NSAIDs are useful in controlling inflammation. These agents are also useful in inflammatory arthritis. Some examples of commonly used NSAIDs include naproxen, ibuprofen, and diclofenac.

Naproxen (Anaprox, Naprelan, Naprosyn)

Naproxen is used for relief of mild to moderate pain. It inhibits inflammatory reactions and pain by decreasing cyclooxygenase activity, which decreases prostaglandin synthesis.

Ibuprofen (Advil, Motrin)

Ibuprofen is commonly used for the treatment of mild to moderately severe pain, if no contraindications exist. It inhibits inflammatory reactions and pain, probably by decreasing activity of cyclooxygenase, which decreases prostaglandin synthesis.

Diclofenac (Voltaren-XR, Cataflam, Zipsor)

Diclofenac is one of a series of phenylacetic acids that has demonstrated anti-inflammatory and analgesic properties in pharmacological studies. It is believed to inhibit the enzyme cyclooxygenase, which is essential in the biosynthesis of prostaglandins.