Henry D. Clarke
Total knee replacement is an excellent treatment option for relieving arthritic knee pain and for improving function. However, it is a technically demanding procedure and the long-term success is dependent on the way the prosthesis is implanted. In this chapter important variables that are associated with achieving optimal results are presented. These include factors related to patient selection, preoperative planning, surgical technique and postoperative management.
Pathogenesis
Etiology
The primary indication for knee replacement is for relief of chronic disabling knee pain owing to arthritis that has failed to respond to nonsurgical treatment regimens. In the United States >90% of the patients who undergo total knee arthroplasty (TKA) have osteoarthritis (OA), with the other main causes including rheumatoid arthritis and posttraumatic arthritis. Risk factors for the development of OA include age, body weight, gender, family history of disease, and prior injuries including meniscectomy and cruciate ligament tears. The increased risk in each of these groups is likely multifactorial with both mechanical and chemically mediated components. The composition of both hyaline cartilage and synovial fluid changes with age in both men and women, but women have higher rates of knee OA and therefore, the role of gonadal hormone levels has been debated. A genetic component has also been implicated, although it is likely that many genes contribute to this risk. Genetic differences in cartilage and subchondral bone constituents appear to be involved in this process, and research is ongoing to identify important markers. Patients with a high body mass index are also at increased risk for the development of OA, and this risk may be owing to factors beyond the simple mechanical trauma produced by the elevated joint forces. Clearly some or all of these risk factors may be identified in any individual, and therefore the relative contribution of each is hard to determine.
Epidemiology
Population studies from Western societies have estimated that approximately 10% to 20% of adult patients older than 35 years of age experience chronic knee pain. Among this group of patients, the prevalence of radiographically identifiable OA has been reported to be between 1% and 70%, depending on the age subgroup, but most studies have suggested a rate of between 1% and 15%. In the United States Medicare population, between 30 and 70 people per 10,000, depending on gender and age subgroup, underwent primary TKA in 2000, with the highest rate of 67.9 per 10,000 noted in women between the ages of 75 and 79 years. During this time period, >350,000 primary TKAs were performed annually in the United States.
Diagnosis
Clinical Features
History
Patients with arthritic knees complain of pain that is exacerbated by weight-bearing activity and relieved by rest. Pain may be isolated to one area of the knee in unicompartmental arthritis or diffuse when multiple compartments are involved. Anterior knee pain that is aggravated by stairs or arising from a sitting position is suggestive of patellofemoral involvement. Secondary symptoms include varying degrees of swelling, buckling or giving way, catching, grinding, and stiffness. In some cases, pain may be referred from other areas of the body to the knee. Other causes should always be considered, especially in cases where the pain is not clearly activity related; radiates to the hip, back, or foot; or is inconsistent with the associated physical exam or radiographic studies. In these circumstances, arthritis of the hip, lumbar radiculopathy, and inflammatory diseases without joint destruction should be considered. It is also important to elicit a history of failed nonoperative and operative treatments including steroid or hyaluronic acid injections, oral anti-inflammatory medication, physical therapy, bracing, arthroscopic debridement, ligament reconstruction, and osteotomy.
P.171
In addition to reliving arthritic pain, a secondary goal for total knee arthroplasty is to improve patient function. Evaluating the degree of disability experienced by a patient can be difficult but is crucially important. Patient expectations and goals must be clearly understood to optimize satisfaction postoperatively. Some patients may simply be hoping to be able to perform activities of daily living without pain, whereas others may be expecting to be able to participate in vigorous sports such as marathon training or basketball. In patients with high expectations, a frank discussion of the goals of knee replacement and the types of activities that can be realistically pursued postoperatively is necessary.
Relative contraindications to total knee replacement include a history of prior infection in the involved knee or active infection in any other location, neuromuscular conditions such as Charcot arthropathy, prior fusion of the involved knee, and a nonfunctional extensor mechanism. Absolute contraindications include active infection in the involved knee or periarticular region, severe peripheral vascular disease, and lack of adequate soft tissue coverage.
It is also important to elicit a history of any systemic conditions that the patient has experienced. Obesity, diabetes, coronary or pulmonary disease, peripheral vascular disease, and immune compromise owing to cancer or HIV all increase the risks associated with joint replacement. Comprehensive consultation with appropriate specialists is critical to optimize results but rarely precludes joint replacement as long as the patient understands the potential risks.
Physical Examination
Preoperative evaluation should determine the presence of prior skin incisions about the knee, the overall clinical alignment of the leg, joint line and peripatellar tenderness, crepitus, whether the collateral ligaments are competent, and whether the varus or valgus deformity is passively correctable to neutral. In cases where the ligaments may be incompetent, a prosthesis with increased femoral-tibial constraint should be available for use. In addition, the passive range of motion, fixed flexion contractures, and extension lags should be noted. Distal pulses, strength and sensation in the extremity should also be evaluated. Finally, as previously noted, absence of significant hip pain with passive motion and adequate hip range of motion should also be verified.
Radiologic Features
Radiographs of the affected knee are adequate for preoperative counseling in most cases. Standard views include a weight-bearing anterior-posterior (AP) view in full extension, a lateral view, and a Merchant view of the patellofemoral joint. In some cases a posterior-anterior (PA) view in 45 degrees of flexion is helpful for demonstrating significant joint space narrowing when the AP standing view shows only minimal changes. The PA flexion view provides a superior view of the contact between the distal-posterior femoral condyles and the tibia, which is an area where significant cartilage wear can occur. In cases where joint space narrowing is unremarkable or where pain is out of proportion to the radiographic evidence, MRI of the knee may identify meniscal pathology or other periarticular pathology such as avascular necrosis, stress fractures, and bone lesions that require alternative treatment. If lumbar or hip pathology is suspected after physical exam, then adequate radiographs of these areas should also be obtained.
Prior to knee replacement a full length, three-joint view of the lower extremity helps with preoperative planning. However, this is probably only mandatory in cases where there is a history of prior fractures or surgery of the ipsilateral extremity, or physical exam suggests unusual extra-articular deformities.
Treatment
Surgical Technique
The key components of the surgical technique for total knee replacement include selecting placement of the skin incision, gaining adequate exposure to the joint, restoring axial alignment of the limb by accurately resecting bone from the femur and tibia, and creation of symmetric flexion and extension gaps with balanced medial and lateral soft tissue tension by releasing the contracted structures. Subtle differences exist depending on whether a posterior cruciate retaining, substituting, or sacrificing prosthesis is implanted, but the broad principles are the same and these are presented in the subsequent sections.
Skin Incision
An anterior midline, vertically oriented skin incision that deviates slightly to the medial side of the tibial tubercle distally is the most utilitarian approach to the knee. Traditionally, incision length was between 15 and 20 cm depending on the size of the patient and surgeon preference. However, with increased emphasis in recent years on reducing incision length and soft tissue dissection in so-called minimally invasive techniques, the incision length has declined and various authors have described the ability to perform TKA through shorter skin incisions.
Prior incisions must be treated with caution as the blood supply to the knee is limited. The vascular supply to the overlying skin is medially biased, and this should be considered in the decision about incision placement when prior incisions exist. In particular, wide scars, lateral incisions, and skin with posttraumatic or postradiation scarring or thinning should be treated with special concern. In general, a single pre-existing transverse incision may be crossed at a right angle with little concern. If a prior anterior incision is present, it should be used unless it lies too far medial or lateral. In circumstances where a prior vertical incision is significantly displaced from the midline, especially with short, well-matured scars, a second vertical midline incision can be made if an adequate skin bridge of about 5 cm can be maintained. However, if multiple vertical or mixed anterior incisions are present, alternative techniques such as tissue expanders or even prophylactic muscle flaps may be required to reduce the risk of postoperative wound-healing problems.
P.172
Exposure
The exposure is largely independent of the skin incision that has been used, although with recent minimally invasive techniques, placement of the short incision over the area where the arthrotomy will be performed is optimal. The most utilitarian approach to the knee joint is via a medial parapatellar arthrotomy that begins 5 to 8 cm proximal to the superior pole of the patella, about 5 mm lateral to the medial border of the quadriceps tendon. The arthrotomy extends distally either around the medial border of the patella or directly over the medial edge of the patellar and then extends along the medial edge of the patellar tendon about 5 to 8 cm distal to the joint line. Next, the anterior horn of the medial meniscus is transected and the medial capsule and periosteum is elevated from the proximal 3 to 4 cm of the medial tibia. The infrapatellar fat pad is resected, and the lateral patellofemoral ligament is divided. If at this stage the patella cannot be subluxated laterally, or everted from the field of view, a quadriceps snip can be performed. Beginning at the apex of the arthrotomy in the quadriceps tendon, the arthrotomy is extended laterally and superiorly at an angle of 45 degrees into the vastus lateralis muscle. In the rare case where this maneuver does not relieve tension on the extensor mechanism and the exposure is still inadequate, a tubercle osteotomy can then be performed and will provide adequate exposure.
The recent trend to minimally invasive surgery has prompted renewed interest in alternative approaches to the anterior knee that include subvastus, mini midvastus, and medial and lateral capsular incisions. Although all of these alternatives are believed to cause less damage to the extensor mechanism and allow quicker functional recovery, few controlled studies exist. Furthermore, the visualization of the knee with any of these exposures is limited, and therefore, they are not suitable for every patient in all surgeons' hands. In particular, patients with heavily muscled thighs, obese patients, and patients with patellar baja or large deformities pose special challenges and may not be amenable to these limited approaches.
|
Figure 25-1 The mechanical axis of the knee should pass through the center of the hip, knee, and talus once the prosthesis has been implanted. Both the femoral and tibial components are oriented perpendicular to the mechanical axis. The femoral component is in approximately 5 to 7 degrees of valgus relative to the anatomic axis of the femoral shaft. |
Bone Resection
Restoration of axial mechanical alignment of the operated leg within a narrow range of ±3 degrees has been demonstrated to be an important determinant for long-term success following TKA. Therefore, bone resection must be performed in an accurate and reproducible way. Orientation of the femoral and tibial components parallel to the mechanical axis of the leg is the goal of the bone resection in TKA (Fig. 25-1). Many instruments have been designed to help the surgeon optimize the bone resection of the distal and posterior femur and proximal tibia. These include both intramedullary and extramedually alignment guides and cutting blocks that are affixed to the bones. Recently computer navigated systems that use either optical or electromagnetic sensors have been developed to aid in this task and have demonstrated more reproducible results than mechanical guides. The specific order of femoral and tibial cuts is irrelevant as these steps are independent in the classic method of bone resection that is favored by many surgeons. It must be recognized that some surgeons favor the use of tensor systems that rely on a tibial cut made perpendicular to the mechanical axis of the tibia to determine the femoral cuts. In this technique, the tibial cut must be made first.
In a knee with varus deformity, the distal femur usually should be cut in 5 to 7 degrees of valgus relative to the femoral shaft or anatomic axis. However, to avoid persistent excessive valgus alignment in a valgus knee, a distal femoral cut of 4 to 5 degrees of valgus relative to the anatomic axis is suggested in these cases. A three-joint view of the limb can facilitate selection of the optimal distal femoral resection by allowing the angle between the anatomic and mechanical axes of the femur to be measured for the specific individual. Other variables such as the placement of the starting hole and fit of the intramedullary alignment guide in the femoral canal can affect the accuracy of the cut and probably have more of an influence on the ultimate resection angle than surgeon choice of 5 or 6 degrees.
|
Figure 25-2 The femoral component is aligned parallel to the transepicondylar axis, which passes through the center of the prominence of the lateral epicondyle and the center of the sulcus of the medial epicondyle. |
The next important step is to accurately size the femur and set the femoral component rotation (Fig. 25-2). This step will determine the anterior femoral and posterior femoral condylar resections. The epicondylar axis has been shown to be the most reliable landmark for determining accurate rotation and is easily identified intraoperatively. If the femoral component is not set parallel to this axis, it
P.173
is difficult to produce a symmetric flexion space. The AP axis, or so-called Whiteside line, is a good secondary reference point that links the center of the intercondylar notch and the center of the femoral trochlea. This axis is usually perpendicular to the epicondylar axis. With the femoral cutting block oriented relative to these landmarks, in most circumstances, more bone will be resected from the posterior medial condyle than the lateral condyle because the epicondylar axis is externally rotated relative to the posterior condylar line. In a varus knee the epicondylar axis is generally externally rotated by about 3 degrees relative to the posterior condylar line, whereas in the valgus knee the epicondylar axis tends to be externally rotated by about 5 degrees.
Next, the tibial resection guide is set to produce a bone cut perpendicular to the mechanical axis. Approximately 9 to 10 mm of bone typically will be resected from the unaffected compartment, i.e., from the lateral side in a varus knee. Once the distal and posterior femoral cuts and tibial cut have been made, a spacer block with an extramedullary guide rod is inserted to evaluate whether the optimal limb alignment has been achieved. If the bone cuts fail to achieve the desired limb alignment, then soft tissue balancing of the medial and lateral structures may be difficult to achieve. Furthermore, as previously noted, detrimental mechanical stresses associated with chronic malalignment can lead to progressive laxity and instability. If overall alignment is acceptable, then the next step is creating balanced and symmetric flexion and extension gaps.
Flexion and Extension Gap Balancing
Soft Tissue Releases.
Evaluation of the soft tissue tension about the knee begins with an examination of the extremity under anesthesia to evaluate the integrity of the collateral soft tissue restraints. If the deformity can be corrected to neutral, a less aggressive soft tissue release should be anticipated than in a knee with a fixed deformity. Once the bone cuts have been performed as noted previously, re-evaluation of the medial and lateral soft tissue tension is performed with a spacer block as previously noted. In addition to bone and cartilage erosion, the development of deformity associated with degenerative arthritis involves the development of contractures of the soft tissue structures on the concave side of the deformity, and eventually, stretching of the structures on the convex side. For example, in the valgus knee, the lateral structures shorten and the medial soft tissues may become attenuated. The goal of soft tissue balancing is to release or lengthen the tight structures to create symmetric, rectangular flexion and extension spaces (Fig. 25-3). Although mild degrees of soft tissue imbalance may be clinically insignificant, it seems prudent to strive for optimal balance. The techniques described below for soft tissue balancing are based on the principles described by Insall.
|
Figure 25-3 Equal and symmetric flexion and extension gaps are created by bone resection and soft tissue releases. |
In the varus knee, the contracted medial structures include the pes anserine tendons, superficial medial collateral ligament (MCL), posteromedial corner including the semimembranosus insertion, and deep MCL. After the standard arthrotomy and exposure has been performed and the bony cuts on the femur and tibia have been completed, the remnants of the cruciate ligaments and menisci should be excised. This is best performed with the knee in flexion, with a lamina spreader providing gentle joint distraction. It is important to remember that the fibers of the deep MCL attach to the peripheral margin of the midbody of the medial meniscus and must not be damaged during meniscal resection. This is most safely accomplished by leaving a thin rim of 1 to 2 mm of peripheral meniscus. At this stage, posterior condylar osteophytes should be removed with an osteotome.
Next, the largest spacer block that will fit in the flexion gap is inserted and stability is assessed. The knee is then extended and the limb alignment is evaluated. If alignment is acceptable, then the medial-lateral balance is assessed. If the medial structures are still tight, as is frequently found in the varus knee, an incremental release of the medial structures is performed to correct the asymmetry of the medial and lateral soft tissue tension. A ¾-inch straight osteotome
P.174
is used to extend the subperiosteal elevation of the distal superficial MCL insertion and deep fascia along the posteromedial border of the tibia. This release may be extended to approximately the level of the middle third of the tibia. In addition, the pes tendons may be released. In some cases, the popliteus tendon may impinge on the posterolateral aspect of the prosthesis, and in these cases of varus deformity it may be released.
Next, the spacer is reinserted and the efficacy of the release is evaluated. In many cases, once the extension space symmetry has been restored by the release, the next thicker spacer block is required. If an imbalance persists, then further subperiosteal elevation of any palpable tight medial bands should be performed distally. In addition, the tibia should be subluxated and externally rotated out from underneath the femur, and in this position a subperiosteal elevation of the semimembranous and posterior capsule from the posteromedial tibia should be completed if not already done.
In the valgus knee, the contracted anatomic structures include the iliotibial band (ITB), lateral collateral ligament (LCL), popliteus tendon, and arcuate ligament/posterolateral capsular complex. If alignment is acceptable when the knee is brought into extension with the spacer block, but the lateral side is tight, then the spacer is removed and laminar spreaders are inserted and gently opened. The lateral soft tissue structures are then released in a graduated fashion using an inside-out technique with the popliteus tendon as a landmark. The arcuate and posterolateral capsular complex are incised horizontally with a number 15 blade at the level of the tibial bone cut. Next, multiple “pie crusting” puncture incisions are made through the ITB and capsule, both at the level of the extension gap and proximal to the joint. Although no specific attempt is made to divide the LCL, it is likely at least partially cut. Once the extension gap appears rectangular, the spacer block is reinserted and the balance re-evaluated. If at this stage the lateral side is still tight, then further pie crusting is performed. In certain cases, the ITB may need to be released entirely from Gerdy's tubercle.
In the valgus knee, the popliteus tendon is preserved, if possible, to act as a lateral stabilizer in flexion and to help prevent rotatory instability. However, in severe valgus knees, typically greater than about 20 degrees, the lateral side may be tight despite the above-noted releases. In these cases, it may be necessary to strip the lateral femoral condyle including the insertion of the popliteus tendon, either sharply or by elevating a wafer of bone from the lateral epicondyle. In these situations, a constrained prosthesis may be required to provide medial and lateral stability. In elderly patients with large valgus deformities, use of a constrained condylar type of prosthesis has been associated with good long-term results despite the theoretical concerns regarding loosening.
Once balanced flexion and extension spaces have been created in either the varus or valgus knee, the knee is assessed to ensure that the size of the overall gaps is equal. The spacer block that allows full extension to be achieved without any tendency to hyperextension is selected, and finally the flexion space must be re-evaluated to ensure that it is symmetric with the extension gap. In cases where the flexion and extension gaps are not equal, further adjustments to the bone resection may be required.
To ensure that symmetry is achieved in the size of the flexion and extension gaps, a comprehensive understanding of the impact of the three basic bony cuts in total knee arthroplasty is required. The proximal tibial cut affects both the flexion and extension gap equally, whereas the distal femoral cut selectively determines the extension gap and the posterior femoral resection affects only the flexion gap (Fig. 25-4). These basic principles provide excellent guidance if asymmetric size gaps are encountered.
|
Figure 25-4 The tibial resection influences both the flexion and extension gaps, whereas the distal femoral cut affects only the extension gap and the posterior femoral resection influences only the flexion gap. |
If the knee is balanced in flexion but tight in extension, with a persistent flexion contracture, an additional 2 mm of femur must be removed as femoral resection selectively changes only the extension gap. In some cases, elevation of the posterior capsule from the femur can correct a slight tendency to residual flexion contractures, especially in the setting of significant preoperative contractures, but bony resection generally provides a more satisfactory result. If the knee is too tight in both flexion and extension to allow insertion of the smallest 10-mm spacer, then additional tibia must be cut as tibial resection changes both the flexion and extension gaps. In the primary setting, with a posterior cruciate retaining implant, it is uncommon to find the knee balanced in extension and too loose in flexion. This may be encountered in a posterior stabilized (PS) knee where release of the PCL may increase the flexion gap more than the extension space. In this setting, resection of additional distal femur and use of a larger polyethylene insert will be the solution. In rare circumstances with either a cruciate-retaining (CR) or PS knee, overresection of the posterior femoral condyles owing to the use of an anterior referencing femoral cutting guide or undersizing of the femoral component may be responsible. In these cases where a CR prosthesis has been used, restoring the posterior condylar offset
P.175
by upsizing the femoral component and using posterior augments can be considered. With a PS knee, additional distal femur can be resected and a larger polyethylene used. This may move the joint line more proximally, but 5 to 8 mm of elevation is well tolerated with a PS knee. In distinction, joint line elevation with a CR knee is less desirable. Finding that the extension gap is balanced but the flexion gap is tight is more likely to occur in a CR knee where the PCL is too tight. In these cases, graduated release of the PCL or increasing the slope of the tibial cut should be used to solve the gap imbalance.
Component Positioning.
Once appropriate extension and flexion gap symmetry has been obtained, the bone surfaces can be prepared for final component positioning. On the femoral side, chamfer cuts must be made as well as a box cut if a PS prosthesis is used. The femoral component should be lateralized on the distal femur, without creating overhang, to optimize patellar tracking. Next, tibial rotation is oriented relative to the junction of the medial and middle thirds of the tibial tubercle. Internal rotation may result in lateral patellar subluxation. Finally, the patellar component is positioned slightly medially and superiorly on the prepared surface, which helps prevent patellar maltracking. The overall composite thickness of the resurfaced patella should restore, or when possible, slightly reduce (1 to 2 mm) the thickness of the native patella. Once these steps have been completed, a reduction using trial components is performed to ensure that appropriate soft tissue balance has been achieved without flexion contracture or hyperextension. If imperfections exist, adjustments are made. Lastly, a “no thumbs” technique is used to evaluate patellar tracking. If the femoral and tibial rotations have been set correctly, the thickness of the patellar has been reproduced, and the other techniques for optimizing patellar tracking have been used, patellar subluxation is uncommon in the varus knee. However, if no technical errors can be identified and maltracking is present, a lateral patellar release should be performed. Once the result with the trial components is acceptable, the surfaces are cleaned and dried and the real components are cemented in place. Once the cement is hard, I routinely release the tourniquet and cauterize any significant bleeding vessels. The joint is irrigated and a deep drain is placed prior to arthrotomy closure, which is performed in extension. After skin closure, a light sterile dressing is used.
Postoperative Management and Rehabilitation
A multimodal approach is used for perioperative pain control. Both nonsteroidal anti-inflammatory medications and narcotic analgesics are given preoperatively in the holding area. Regional blocks including femoral and sciatic nerve blocks are performed preoperatively and are continued postoperatively for 24 to 48 hours. In conjunction with the regional blocks, intravenous narcotics are administered via a patient-controlled analgesia device for breakthrough pain during the first 24 hours. Patients are then switched to oral narcotics for pain control. Passive and active ranges of motion are begun on postoperative day 1 and are advanced as tolerated; the importance of active extension is emphasized to the patient. Ambulation with weight bearing as tolerated is also begun on postoperative day 1, without limitation. Early goals include independent transfers, walking as tolerated, and active motion from full extension to 90 degrees of flexion. Other important perioperative interventions include the use of prophylactic antibiotics given within an hour of the incision and continued for 24 hours postoperatively, and deep vein thrombosis prophylaxis. A multimodal approach to DVT prophylaxis is also used, including the use of thigh-high compression stockings, mechanical sequential compression devices, and low-molecular-weight heparin or adjusted dose Coumadin. The use of continuous passive motion machines is controversial, and there are studies that both support and refute its efficacy.
Results
During the past two decades, the results of total knee replacement have been proven both consistent and durable. Indeed, long-term survivorship has been reported from independent centers to be >90% to 95% at 10 years or greater. In these studies, various prosthesis designs have demonstrated excellent results in both young and old adults. Despite these highly reproducible outcomes, failures do occur. Infection, mechanical failure, periprosthetic fracture, aseptic loosening, polyethylene wear, and instability are the most common modes of failure. Although some of these problems may be unavoidable, long-term success has clearly been noted to be related to patient characteristics and the accuracy with which the prosthesis is implanted. Therefore, both careful preoperative evaluation and optimal surgical technique should be used and remain within the control of the orthopaedic surgeon.
Suggested Readings
Berger RA, Rubash HE, Seel MJ, et al. Determining the rotational alignment of the femoral component in total knee arthroplasty using the epicondylar axis. Clin Orthop. 1993;286:40–47.
Clarke HD, Scuderi GR. Correction of valgus deformity in total knee arthroplasty with the pie-crust technique of lateral soft-tissue releases. J Knee Surg. 2004;17(3):157–166.
Clarke HD, Scuderi, GR. Revision total knee arthroplasty: planning, management, controversies, and surgical approaches. Instr Course Lect. 2001;50:359–365.
Insall JN, Easley ME. Surgical techniques and instrumentation in total knee arthroplasty. In: Insall JN, Scott WN, eds. Surgery of The Knee. Vol 2. 3rd ed. Philadelphia: Churchill Livingstone; 2001:1553–1628.
Keating EM, Meding JB, Faris PM, et al. Long-term followup of nonmodular total knee replacements. Clin Orthop Relat Res. 2002;404:34–39.
Kelly MA, Clarke HD. Long-term results of posterior cruciate-substituting total knee arthroplasty. Clin Orthop. 2002;404:51–57.
Lonner JH, Siliski JM, Scott RD. Prodromes of failure in total knee arthroplasty. J Arthroplasty. 1999;14:488–492.
Mahomed NN, Barrett J, Katz JN, et al. Epidemiology of total knee replacement in the United States Medicare population. J Bone Joint Surg Am. 2005;87:1222–1228.
P.176
Malkani AL, Rand JA, Bryan RS, et al. Total knee arthroplasty with the kinematic condylar prosthesis. A ten-year follow-up study. J Bone Joint Surg Am. 1995;77:423–431.
Ritter MA, Faris PM, Keating EM, et al. Postoperative alignment of total knee replacement. Its effect on survival. Clin Orthop. 1994;299:153–156.
Scott RD, Thornhill TS. Posterior cruciate supplementing total knee replacement using conforming inserts and cruciate recession. Effect on range of motion and radiolucent lines. Clin Orthop. 1994;309:146–149
Sowers, MF. Epidemiology of risk factors for osteoarthritis: systemic factors. Curr Op Rheumatol. 2001,13:447–451.