Jeffrey S. Zarin
Andrew R. Noble
Wolfgang Fitz
In the United States, over 300,000 total knee arthroplasties are performed each year. Total knee arthroplasty (TKA) is one of the most successful procedures in orthopaedic surgery, and there are excellent reported long-term results with survivorship rates of >90% at 15 years.1,2,3 In recent years surgical techniques have been changed and new technologies have been introduced for TKA. Patients are now more informed and are requesting newer technologies, better implants, less pain, less blood loss, and quicker recovery from a joint replacement. The introduction of minimally invasive techniques, preemptive analgesia, progressive rehabilitation, computer-assisted surgery, and new materials not only have changed the daily practice for orthopaedic surgeons but also have added new challenges. Beyond the scope of these advances, surgeons must keep in mind that TKA is a major surgery with associated morbidity. This chapter will focus on the postoperative complications including traumatic periprosthetic injuries, the pitfalls of minimally invasive surgery, wound healing problems, nerve injury, and postoperative infection. Also, issues associated with stiffness, tissue balancing, and instability will be addressed.
Morbidity and Mortality of Tka
The overall estimated mortality for total knee arthroplasty during the first 90 days is 0.2% to 0.7%. Increased risk is associated with advanced age, comorbidities, and revision procedures. In a study of >3,000 consecutive TKAs performed by one surgeon, the overall mortality rate was 0.46% during the first 90 days in patients with an average age of 70 years.4 Gill et al.4 reported a risk of mortality 16 times higher in patients with cardiac comorbidities such as previous myocardial infarction, ischemic heart disease, and cardiac failure compared with those with no comorbidity. Patients older than 85 years of age had a 14-times increase in the chance of death when compared with patients younger than 85 years of age, with a reported rate of 4.65%. The mortality rate in the Medicare population undergoing primary TKA is reported as 0.6% to 0.7% during the first 90 days in two studies.5,6 The overall morbidity rates in >80,000 patients during the first 90 days after primary TKA identified in a Medicare population were the following: acute myocardial infarction, 0.8%; pulmonary embolism, 0.8%; pneumonia requiring hospitalization, 1.4%; and infection requiring irrigation and debridement, 0.4%.5,6
Surgeon and Hospital Volume
The relationship of surgeon volume to patient outcomes has become a topic of increasing interest. Two recent studies have reported lower mortality and morbidity rates associated with surgeons and hospitals performing a larger volume of TKAs. Katz et al.5 identified a 30% reduction in mortality rate for patients receiving a TKA in hospitals that perform >25 of these procedures per year. Surgeons performing >50 procedures per year had a 40% lower risk for deep wound infection compared with surgeons performing <12 per year. A steady decline of deep infection was independently related to hospital volume as well, with a reported 40% reduction for hospitals performing >200 cases per year versus those doing <25 per year. The risk of pneumonia also diminished independently for surgeons and hospitals performing >12 and 25 cases, respectively.
Periprosthetic Fractures
Periprosthetic fractures of the femur, tibia, or patella are rare after total knee replacement. The reported prevalence for distal femoral fractures ranges from 0.3% to 2.8%7 and for tibial fractures from 0.4% to 1.7%.8 Patellar fractures occurred in 0.05% when unresurfaced9and ≤21% with resurfacing.10
P.178
Distal Femoral Fractures
Supracondylar femur fractures are most frequently traumatic, with a higher incidence seen in patients with osteopenia. Femoral notching may increase the incidence of periprosthetic fractures of the distal femur when both medial and lateral cortices are notched. Nondisplaced fractures with well-fixed implants can be treated by nonoperative intervention with a high success rate. Surgical intervention is required in the setting of displaced supracondylar fractures, and the method of fixation is determined by implant stability. Supracondylar fractures associated with well-fixed femoral components can be treated by several techniques including retrograde nails, blade plates, condylar screw plates, condylar buttress plates, and locked condylar plates.
Retrograde nailing and more recently less-invasive condylar locking plates have become standard treatment methods owing to the preservation of fracture hematoma and minimal soft tissue dissection. Whether the fracture is suitable for a retrograde nail is determined by the length of the distal bone fragment from the fracture to the intercondylar notch. Adequate bone length of the distal fragment is needed for placement of the two distal locking screws. If the most distal aspect of the nail protrudes into the notch, some surgeons have successfully removed this portion after inserting the interlocking screws.11 The design of the femoral component must also be considered because posterior stabilized systems may preclude insertion of the retrograde nails through a solid cam and post mechanism. Locked plates inserted with minimal soft tissue disruption offer many advantages over retrograde nailing, including rigid fixation with locked screws, ability to combine with posterior stabilized systems, and potentially better fixation in osteopenic patients.12
Loose femoral components in combination with supracondylar fractures require a different treatment approach. In certain cases, the periprosthetic fracture can be addressed first and allowed to heal prior to revision of the loose femoral implant. Postponing component revision until fracture healing offers several advantages including less bone loss, ease of revision TKA, reduced need for cortical strut allograft, and less need for augments, wedges, stems, and constrained or hinged prostheses.8 Combined fixation of the fracture with revision knee arthroplasty is a technically demanding procedure that may require extensive allografts and a hinged prosthesis or oncologic distal femoral replacement prosthesis. Principles include restoration of the joint line, preservation of fixed components, and proper femoral rotation based on a rectangular flexion gap with the tibial component. Bulk allograft may be necessary to restore condylar bone loss. The use of extensive bone cement at the fracture site is discouraged because of risk of nonunion.
Tibial Fractures
Undisplaced or reducible tibial fractures that remain in a stable anatomic position are amenable to nonoperative treatment. Displaced and unstable fracture patterns associated with well-fixed total knee components usually are treated with open reduction and internal fixation with buttress plates or locking plates. Revision total knee replacement is indicated when the fracture involves the tibial component or when the implant is loose. Long-stemmed tibial components should be used to bypass the fracture site and are often secured with a hybrid cement technique. Additional plating or use of bulk allograft may be required based on the fracture pattern and bone loss.
Patellar Fractures
Many factors predispose to patella fractures in TKA. The risk for fracture in nonresurfaced patellae is minimal. Extensive resection with a patella thickness of <15 mm can predispose to fracture.13 A three-peg design has reduced patellar strain and has a decreased likelihood of fracture compared with a larger single peg. Several other risk factors for patellar fracture have been identified and include overstuffing of the femoropatellar joint, use of oversized femoral components, component malrotation, and placement of the femoral component in too much flexion.14
Disruption of the patellar blood supply is another important factor leading to avascular necrosis (AVN) and eventual patellar fracture after total knee replacement. The patellar blood supply may be compromised when a median parapatellar approach is combined with a lateral release. Scuderi et al.15 demonstrated a 56.4% incidence of reduced blood flow to the patella when a lateral release was performed following a parapatellar approach. However, when a medial subvastus approach is used, there is less risk for AVN when combined with a lateral release because the superior geniculate artery is preserved. No data are available to show that the decreased exposure and reduced soft tissue violation of minimally invasive surgery has an effect on patellar blood supply and associated fractures.
Ortiguera and Berry10 classified patellar fractures based on fixation of the patellar component, integrity of the extensor mechanism, and quality of the residual patellar bone stock. The fractures are classified as type I with a stable implant and an intact extensor mechanism, type II with disruption of the extensor mechanism, and type III with a loose patellar component and reasonable bone stock (>10 mm thickness, IIIA) or poor bone stock (<10 mm thickness or comminution prohibiting fixation or resurfacing, IIIB). In this study comprising 78 patella fractures, about half were classified as type I and were treated successfully with observation or immobilization.
Disruption of the extensor mechanism typically is treated with surgical intervention. However, type II fractures were associated with a high complication rate of 50% and a reoperation rate of 42%. Open reduction internal fixation was rarely successful owing to a very thin and small piece of bone. Other surgical options included partial or total patellectomy with repair and advancement of the extensor mechanism.Figure 26-1 shows an open reduction internal fixation of a type II fracture with complete rupture of the extensor mechanism. Intraoperatively, it was felt that the remaining distal pole of the patella was large enough for fixation; it ultimately healed without an extension lag or quadriceps weakness (Figs. 26-2, 26-3, 26-4).
Failure of extensor mechanism repair typically is salvaged with an allograft reconstruction consisting of tibial tubercle, patellar tendon, patella, and quadriceps tendon that was first described by Emerson et al.16 Nazarian and Booth17 modified this technique by tightly tensioning the
P.179
repair in full extension and reported improved early results. Burnett et al.18 reported a series of 20 consecutive reconstructions with one group having minimal tension in extension whereas the second group was tightly tensioned intraoperatively. Loosely tensioned allografts resulted in persistent extensor lag and clinical failure. The tightly tensioned reconstructions were all clinically successful with an average postoperative extensor lag of 4.3 degrees.
|
Figure 26-1 Fracture of the inferior patella pole with complete extensor mechanism disruption. |
|
Figure 26-2 Lateral view of left knee with healed repair of patella fracture and extensor mechanism without functional deficit. |
|
Figure 26-3 Skyline view of repaired and healed patella pole fracture and ruptured extensor mechanism. Au: Is expansion of s/p correct? |
|
Figure 26-4 Anteroposterior view of bilateral total knee arthroplasties and healed left patella fracture. |
Wound Healing Problems
Early wound healing problems after total knee arthroplasty occur infrequently but should be suspected in higher-risk patients who are immunosuppressed, malnourished, taking steroids, or have diabetes or rheumatoid arthritis, as well as those with a history of multiple surgeries or prior infection in the operative knee.
Small amounts of wound drainage that lightly stain dressings may commonly be seen in the first 3 to 4 days after
P.180
surgery. The treatment of postoperative drainage is not clearly presented in the orthopaedic literature but is based on sound clinical judgment. Drainage should be more concerning to the surgeon when it continues after 5 days and if it is associated with diffuse erythema, purulence, or profuse volume. Persistent drainage, particularly of serosanguineous character, usually is an indication for aspiration and consideration of open irrigation and debridement.
Morbidly obese patients undergoing TKA are at increased risk for subcutaneous fat necrosis and potential wound drainage. Application of an incisional vacuum sponge has been introduced and promoted by orthopaedic trauma surgeons to potentially reduce early drainage and wound breakdown in the morbidly obese. After skin closure, an incisional vac with a ½-inch-wide strip of sponge is directly applied to a nonadhesive dressing over the closed incision. After 2 to 3 days, the vac dressing and sponge are removed and replaced with a dry dressing (MB Harris, personal communication, 2005).
A suture abscess may present as an infection but is more often a granulomatous reaction to the suture material. The perplexing diagnosis of suture granuloma is more commonly discussed in the general surgery literature, with only a handful of orthopaedic cases being reported. Three cases of culture-negative granulomatous reactions to Vicryl suture were reported within 9 weeks after total hip arthroplasty by Sayegh et al.19 All cases were successfully treated with excision of the affected tissue, debridement of the joint capsule, and extensive wound lavage. The implants were left in place and the patients were treated with antibiotics pending the negative culture reports, at which time the antibiotics were discontinued. Regarding suture abscesses in total knee replacement, we recommend removal of visible sutures associated with superficial reactions and more formal debridement and antibiotic coverage for deeper cases.
Early postoperative bleeding into a drain is expected but should be more closely observed if profuse and continuous. Temporary immobilization of the knee and avoidance of early motion can often result in spontaneous resolution. However, significant intra-articular hematoma with incisional leakage and excessive soft tissue expansion with impending skin necrosis are indications for prompt formal surgical evacuation with hemostasis. Evacuating the hematoma by squeezing the wound or probing are strongly discouraged because of the potential for retrograde contamination.20
Successful treatment of skin necrosis depends on early recognition and is based on the size, depth, and location of the defect. Superficial skin necrosis <4 cm2 with remaining coverage of bone and tendon may be treated with wet to dry dressing changes or a wound vac. Close observation is imperative to avoid deeper penetration and possible contamination of the prosthesis. An early plastic surgical consult is strongly recommended. For deeper and larger defects of >4 cm2, a plastic surgeon should plan for local flap coverage. Ries23 described the use of a medial gastrocnemius flap or latissimus free flap for defects over the patellar tendon and tibial tubercle. Five of the six patients who underwent flap coverage required two-stage revision total knee replacement. Additional adjunctive treatment measures include immobilization as well as appropriate antibiotic therapy with infectious disease consultation.
Deep Infection
Incidence and Risk Factors
All operative procedures are susceptible to bacterial contamination, and the presence of biomaterials places patients undergoing joint replacement at increased risk for the development of deep infection. The incidence of deep infection has been reported to range from 1% to 2.5% in primary TKA and approaches 5.6% in revision TKA. Factors leading to deep infection must be considered with respect to the microbiologic characteristics of the infecting organism, the host, wound, and operative technique.24
Biomaterials have an increased susceptibility to bacterial contamination because of a self-perpetuating enlarging immunoincompetent fibroinflammatory zone that develops around the implants.25 Bacteria may adhere to the implant based on the surface characteristics and the intrinsic properties of the bacteria. Once adherent, bacteria can encase themselves in a hydrated biofilm matrix of polysaccharide and protein. Sessile, biofilm-encased bacteria are less susceptible to antibiotics than free-floating bacteria.25 This quality of deep bacterial infection of TKA underlies its difficulty in eradication without complete hardware removal.
Patient-specific factors contribute to elevated risk for deep infection as well. Patients with decreased immunity, prior history of deep infection, and higher contamination loads have incidence rates of deep infection between 3% and 10%.25 Patients with decreased immunity include those with rheumatoid arthritis, diabetes mellitus, organ transplantation, obesity, HIV, poor nutritional status, and hemophilia. Patients with increased contamination loads include those undergoing revision total joint replacement and those with surgical duration >2.5 hours. There is evidence that preoperative nasal screening, topical treatment, and specific perioperative antibiotic prophylaxis in combination with vancomycin reduces the incidence of MRSA infection in orthopaedic operated patients to almost zero.27,28
Diagnosis
The key to successful treatment of deep infection is early and accurate diagnosis. Classic clinical presentation of an infected TKA is characterized by increasing persistent pain, warmth, effusion, and less frequently, erythema. Patients with prolonged postoperative pain should be suspected to be infected and should be evaluated for infection. Aspiration of a suspected infected TKA should be performed early and before the first administration of antibiotics.
Repetition of aspiration may increase sensitivity and specificity, as well as increase the chance of identification of the infectious organism with susceptibilities.30 In a two-stage reconstruction of an infected TKA after hardware removal, followed by a 6-week period of intravenous antibiotic therapy, antibiotic therapy should be discontinued for a ≥10 ten days prior to aspiration. Aspiration has been shown to have a 74% positive predictive value and 94% negative
P.181
predictive value, although rates have been identified in studies of ≤100% sensitivity and specificity.24 Newer techniques of polymerase chain reaction (PCR) may increase the utility of aspiration in sensitivity but may be associated with an increased false-positive rate. One recent study evaluated the differential gene expression by white blood cells, using a commercially available gene chip. They identified expression of genes from neutrophils present at the site of infection that was different than that expressed at a site of aseptic inflammation. These findings may lead to potential future simple lab tests that can distinguish the causes of inflammation in total joint arthroplasty.29
Blood tests should include erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) level. However, the sensitivity and specificity of ESR has been reported as low as 60% in one series.24 Therefore, blood tests may be useful for screening but should not be used for the definitive diagnosis of deep infection. Radionucleotide studies, such as indium-labeled leukocyte scans, have been used. Sensitivity and specificity have been reported between 80% and 94%. Increased scan activity can be present in ≤90% of tibial and 65% of femoral components ≥1 year after implantation.24,30
Classification and Treatment
The timing of onset from the index procedure defines the classification system of deep infection and can help guide appropriate management. Deep infections have been classified as those with positive intraoperative cultures, early postoperative infection, acute hematogenous infection, and late chronic infection.
Positive intraoperative cultures may occur in the setting of revision TKA for presumed aseptic loosening. Culture results must be interpreted in conjunction with preoperative examination findings and the overall clinical scenario. Multiple intraoperative cultures can help resolve the dilemma of whether a positive intraoperative culture represents contamination or infection. Greater than two of five positive intraoperative cultures can indicate true infection. If felt to be a true-positive, treatment with a 6-week course of suppressive antibiotics can be curative in 90% of patients.30 This approach is similar to direct-exchange arthroplasty for low-grade infections.
Early postoperative infection occurs within 1 month after implantation. Acute hematogenous infection occurs with seeding of the joint from another primary site of infection, such as urinary tract infection or pneumonia. Invasive procedures leading to transient bacteremia, such as colonoscopy and dental procedures, may contribute to deep joint infection. Presentation is with local inflammation of acute onset and systemic toxicity. Prompt surgical intervention is mandatory, as delays of >2 weeks are associated with decreased rates of implant salvage. A success rate of 60% to 80% has been found with treatment with retention of the implants and multiple debridements. In a study of 24 patients with infected TKA presenting within 30 days of the index procedure or with <30 days of symptoms (acute hematogenous group), Mont et al.31 reported that the implants were successfully retained in 83% of patients after one to three procedures. On the other hand, Deirmengian et al.32 reported on a series of 31 TKA patients with disappointing results of infections with Staphylococcus aureus. Only one was treated successfully with early debridement, liner exchange, and retention of the implants.
Late chronic infection occurs >1 month after the index TKA and involves extension of the infection through the capsule, with or without sinus formation. Onset is more gradual, with slow deterioration of function and increase in pain. The treatment of late chronic infections has received much attention in the literature of the past 30 years. Single-stage revision in the presence of low-grade organisms has been reported. However, most reports favor a two-stage approach with placement of a temporary antibiotic-impregnated cement spacer after a thorough debridement of the knee joint.33 The current recommended dosage is ≥3.6 grams of antibiotics per 40 g of acrylic cement for effective elution kinetics and sustained therapeutic levels of antibiotics.34 We currently use 2 g of vancomycin and 3.6 grams of tobramycin per 40 g of acrylic cement. Premixed antibiotic cements with 1 gram of gentamicin are not recommended for the treatment of deep infection, and an inadequate dosage of antibiotics within bone cement has been described as a cause of treatment failure.35 Intravenous antibiotics appropriate for the infecting organism are administered for 6 weeks, followed by a second-stage implantation of a permanent prosthesis using low dose antibiotic-impregnated bone cement. Success rates have been identified for two-stage replantation of 80% to 93% when using an antibiotic cement spacer.36,37,38,39,40,41
Antibiotic spacer blocks used during the first stage of the two-stage treatment algorithm lead to knee stiffness and may compromise bone stock. Multiple studies have examined the use of an articulated spacer technique. A comparison of static with articulating spacers identified improved preservation of bone stock, increased ease of exposure during replantation, and no apparent increase in reinfection when using articulated spacers.38 One recent study showed the average range of motion with the articulated spacer was 110 degrees, which was not significantly different than the motion after replantation.33 Success rates for eradication of infection with the PROSTALAC spacer were found to be 91%.45 Multiple articulating designs have been described, including all-antibiotic–laden cement, cement and metal composites, and replacement of the original components after autoclave sterilization and loose antibiotic cement technique.33,3637433940 Many variations of spacer design have been described as well, ranging from ball-and-socket type molding, commercially available PROSTALAC designs, and metal-polyethylene-cement composites.45,44 Most involve the intraoperative production of separate femoral and tibial casted or sculpted cement spacers that mimic the design of the metal implants and allow for motion at the cement/cement interface.38,433940 All designs show success rates ≥90% for infection eradication and improved patient function with the articulating spacer.
Stiffness
Definition and Incidence
A severely stiff knee after TKA is an uncommon but disappointing and disabling occurrence. Gait studies have suggested increased difficulty of walking occurs with increasing flexion contracture and that flexion of 67 degrees is required
P.182
for normal gait.46 Stair climbing requires 83 degrees of flexion, rising from a seated position requires 93 degrees, and tying shoelaces requires 106 degrees.47
The overall incidence varies based on the definition criteria used. Severe postoperative stiffness has been defined in the literature as flexion <75 degrees and/or the presence of a knee flexion contracture ≥15 degrees.46,48 However, others have also suggested that an arc of motion <70 degrees or a flexion contracture >20 degrees with a total range of motion <45 degrees constitutes postoperative stiffness. Prior studies have indicated an incidence of stiffness as high as 12%. Two institutions recently found the incidence of stiffness after TKA to be from 1.3% to 3.7%, based on large consecutive series of >1,000 primary TKAs in both studies.46,4
Etiologic Factors
Preoperative, intraoperative, and postoperative events can all contribute to a stiff knee after TKA. The range of motion before the index arthroplasty is the most common preoperative predictor of decreased motion after TKA.46,47 This preoperative stiffness can occur from extensor mechanism or capsular contractures. Although these structures may be released during the index procedure, their elasticity may be restricted owing to chronic fibrosis.49 Body habitus may also decrease postoperative motion. Obese patients with short stature have earlier impingement of posterior soft tissues, decreasing total flexion.51 Other patient factors such as posttraumatic arthritis, juvenile rheumatoid arthritis, ankylosing spondylitis, and keloid formation may increase the risk for stiffness.49,51 Patient noncompliance with postoperative rehabilitation protocols often results in suboptimal knee motion.51 Whether minimally invasive techniques and preemptive analgesia reduce the incidence of stiffness remains unclear at this time.
Technical aspects of the index TKA may be intrinsic to postoperative stiffness. These may include overstuffing of the patellofemoral articulation by oversizing the femoral component or increasing patellar thickness. A recent study found that on average, passive knee flexion decreased 3 degrees for every 2-mm increment of patellar thickness.50 Patella height should be restored and not altered.51,52 Appropriate balancing of the flexion and extension gaps is essential. Gap balancing techniques with spacer blocks or tensiometers help to assess and match intraoperatively composite implant thickness. Femoral or tibial malrotation can be avoided by using the gap balancing technique, in minimally invasive TKA, because anatomic landmarks are not very reliable (Fig. 26-5). Postoperatively, the patellar axis should be parallel to the transepicondylar axis and the tibial axis. Skyline views in 50 to 70 degrees can demonstrate appropriate alignment (Fig. 26-6). An excessively tight flexion and/or extension gap, a tight posterior cruciate ligament (PCL), and femoral and/or tibial malrotation with limited bearing excursion are associated with highly conforming prosthetic designs.52
A tight flexion gap is a common error resulting in decreased flexion51 and can be avoided with the use of spacer blocks. If the posterior femoral condylar bone resection is less than the thickness of the posterior condyles of the femoral component, the flexion gap will be decreased.52 In this scenario, the extension gap will be larger than the flexion gap. Erroneous selection of an increased tibial polyethylene thickness to balance the extension gap will further limit flexion. Also, positioning the femoral component too posteriorly, in excessive malrotation in the coronal plane, or placing a component with a larger anteroposterior dimension than the patient's anatomy will result in a tight flexion gap.51 Failure to remove posterior osteophytes sufficiently can block the full sagittal excursion of the tibial polyethylene and prevent full flexion. These osteophytes can also tense the posterior capsule in extension, causing a paradoxic block to full extension as well.46 Decreased extension may result if the distal femoral resection is too distal, particularly in the setting of a pre-existing flexion contracture. A recent study showed that an average value of 9 degrees of femoral contracture is corrected for every 2 mm of distal femoral resection.54
|
Figure 26-5 Balancing gap technique: Medial and lateral soft tissues are tensioned equally and the femoral size determined (anterior referencing). This creates an equal flexion and extension gap without the use of anatomic landmarks. |
|
Figure 26-6 Skyline view demonstrating patellar axis, transepicondylar axis, and tibial long axis being parallel. |
P.183
Femoral and tibial component malposition in the sagittal plane can lead to stiffness postoperatively.53 The femoral component should optimally be at right angles to the anatomic axis of the femur in the sagittal plane. A hyperflexed component can lead to early cam-post impingement and loss of extension in implants. Figure 26-7 shows a lateral knee radiograph of a PCL-retaining TKA, the femoral component in about 15 degrees of flexion and the tibial component with a slope of 15 degrees. This can be tolerated with a PCL-retaining design, but a PS design would lead to peg impingement.
|
Figure 26-7 Lateral view of a total knee replacement showing a PCL-retaining TKA with about 15 degrees of femoral component flexion and a slope of approximately 15 degrees. In a posterior substituting (PS) design, peg impingement is model specific, but may occur as soon as the combined added angle totals 10 degrees. |
A hyperextended component can limit flexion and increase the risks associated with anterior femoral cortical notching. It has been suggested that tibial slope in the sagittal plane should equal the patient's bony anatomy preoperatively. An up-sloped tibial cut (i.e., higher posterior than anterior) will lead to a decreased posterior joint space and decreased flexion. Increased down-slope will increase flexion but my lead to anterior tibial translation and early posterior polyethylene wear. Sagittal plane tibial component balance is more critical in PCL-retaining knees, whereas a flat tibial slope is more appropriate in PCL-substituting knees that depend on the cam and post mechanism for sagittal plane behavior of the prosthesis.
Limited knee flexion may result from imbalance of the posterior cruciate ligament in PCL-retaining knee designs. The PCL may result in overtightening in flexion, and imbalance of the flexion and extension gaps will lead to stiffness. Paradoxically, a lax PCL leading to flexion instability may also lead to stiffness, as anterior femoral translation occurs with increasing knee flexion. This can induce earlier posterior impingement and extensor mechanism tightening, decreasing ultimate flexion.51 Excessive elevation of the joint line with a cruciate retaining implant may lead to patella infera, which has been associated with patellar pain and limited motion.47,49 Joint line elevation of 3 to 10 mm can substantially increase PCL tension and limit flexion.
Treatment and Outcomes
Postoperative stiffness is best managed by prevention. Preoperative patient education, appropriate postoperative analgesia, and aggressive postoperative rehabilitation help to maximize postoperative motion and function. Continuous passive motion (CPM) machines have been useful adjuncts in the immediate postoperative period, but several studies indicate no significant benefit at 1 year after TKA.51,49
The timing of surgical intervention in the setting of a stiff knee replacement remains controversial. Closed manipulation under anesthesia has been shown to be effective when performed within 6 to 12 weeks after primary TKA. Surgical options after 3 months include arthroscopic arthrolysis for focal adhesions, open arthrolysis for general arthrofibrosis, and only if necessary, component revision.49 Reports of series with arthroscopic arthrolysis and PCL release with a manipulation showed an improvement in only 43% of knees, whereas another group showed an average increase of motion by 30.6 degrees.49,55 Open arthrolysis with radical scar excision and ligamentous rebalancing has shown some promise. A “pie crust” quadricepsplasty followed by a gradual manipulation has been recommended.51 Others suggest a quadriceps snip at the time of exposure with similar benefits.49 A recent study combining aggressive arthrolysis with a customized rehabilitation protocol showed a mean increase in range of motion from 63 degrees to 94 degrees in 94% of knees. However, a flexion contracture averaging 9 degrees remained in 39% of the patients. Sixty-seven percent of the patients had Knee Society scores of good or excellent, with improvement from 34 to 77 points.46
Revision total knee replacement is indicated for situations in which an identifiable, intrinsic problem is associated with stiffness. These include situations as discussed above, such as component malposition, incorrect sizing, joint line displacement, inadequate bone resection, and improper soft tissue balancing. One recent study evaluating patients with revision of femoral and tibial components in the setting of stiffness showed Knee Society scores improved from 38 to 87 and arc of motion improved from 55 degrees to 82 degrees in 93% of knees.48Another report suggested less promising results, with only 10 of 15 patients satisfied
P.184
with outcomes, Knee Society scores from 28 to 65, and an increase of arc of motion from 40 degrees to 73 degrees.56 Successful revision of a stiff knee involves identification of extrinsic sources of stiffness that are uncorrectable, such as ipsilateral hip arthrodesis, neurologic disorders, longstanding extrinsic muscle tightness, and systemic inflammatory conditions. Identification of the cause of failure preoperatively or intraoperatively and assessing correction of the problem after placement of the new prosthesis is associated with the best results.52
Instability
Incidence and Risk Factors
Instability after total knee arthroplasty is one of the most common causes of aseptic failure. Although reported incidence rates range from 1% to 2% in all primary total knee replacements, symptomatic instability may account for ≤10% to 20% of patients undergoing revision surgery. Instability may present as medial-lateral instability or flexion instability.
Certain situations increase the risk of an unstable total knee replacement. Greater preoperative deformity requiring large surgical correction and aggressive ligament releases may lead to difficulties with obtaining stability.58 One study of patients with preoperative valgus deformity averaging >10 degrees noted that 17% of patients had instability postoperatively. In addition, the patients with postoperative knee instability had significantly higher postoperative knee pain than those with stable knees.2 Increased strain on the ligaments of the knee may occur with conditions that alter the mechanics of the knee during gait. Neuromuscular pathology such as quadriceps weakness or hip abductor weakness may lead to increased medial forces at the knee, leading to ligamentous laxity and instability. Valgus forces at the knee may be increased by mechanical instability at the ankle with posterior tibialis tendon rupture or at the hip with a valgus alignment of an ipsilateral hip arthroplasty.58
Obese patients have increased risk of iatrogenic collateral ligament damage owing to difficult surgical exposure. The use of minimally invasive instrumentation may ease implantation in these patients. Assessment of component position and appreciation of ligament balance can be more difficult with the increased soft tissues and weight of a large limb. Increased thigh circumference will cause a wide-based gait, which increases stresses on the medial collateral ligament. Any or all of these factors may contribute to postoperative instability in obese patients undergoing TKA.
Axial Instability
Varus-valgus instability is the most common and classic type of instability pattern. This type of instability may result from collateral ligament imbalance or failure, incomplete correction of preoperative deformity, and component malalignment and/or failure.
Inadequate or overrelease of contracted collateral ligaments when balancing soft tissues for a fixed axial deformity causes an asymmetric extension gap, leading to medial-lateral instability. Inadvertent damage to the medial collateral ligament (MCL) also leads to instability in extension. Care must always be taken to protect the MCL when performing the medial proximal tibial cut as well as the medial posterior femoral condylar cut. In this setting, medial collateral ligament advancement or reconstruction alone with postoperative bracing has been recommended. However, more predictable stability can be attained by combined repair of the MCL and use of a constrained implant.
Incomplete correction of varus or valgus deformities may lead to axial instability because of imbalance between medial and lateral ligaments and soft tissues. For example, an uncorrected varus deformity will produce a lax lateral sleeve and tight medial sleeve, causing a varus thrust during ambulation. Patients with medial-lateral laxity may compensate by walking with a stiff-legged gait to avoid the pain associated with a thrust or sensation of buckling of the knee.59 Reconstruction in this situation should be directed toward re-establishing the joint line and appropriate tension in the soft tissue envelope. Asymmetric instability resulting from improper bone cuts or bone loss often requires the use of modular augments or structural bone grafts.
Symmetric varus-valgus instability may be the result of overresection of the distal femur, component loosening, and soft tissue laxity of the medial and lateral collateral ligaments. An overresected distal femur will lead to a larger extension gap than flexion gap. Choosing a thin polyethylene component that fills only the flexion space will lead to a loose extension gap and associated medial-lateral instability as well as genu recurvatum during gait. Loosening of the femoral or tibial component will present as apparent instability on exam. The loose component will tilt with stress, giving the appearance of an unstable opening joint.59 Global soft tissue laxity may occur in patients with connective tissue disorders such as rheumatoid arthritis or Ehlers-Danlos syndrome. This can result in persistent laxity and instability if not recognized at the time of primary TKA.
Flexion Instability
Mismatch of the flexion and extension gap can lead to flexion instability. This may occur with overresection of the posterior femoral condyles, undersizing the femoral component, and excessive tibial slope.59 All of these causes lead to a flexion gap that is larger than the extension gap. If a thin polyethylene insert is chosen that fills only the extension gap, flexion instability will result. Patients present with recurrent effusions and a sense of instability without buckling. They will often mistrust the stability of their knee when descending stairs, and there is often associated start-up pain. Posterior translation of the tibia in flexion leads to areas of soft tissue tenderness anteriorly and can be exacerbated by weakness of the extensor complex.59,60 This could be prevented by using the gap balancing technique: First the extension gap is balanced and the correct insert thickness selected using a spacer block or a tensiometer. Second, femoral rotation is not based on anatomic landmarks since it is sometimes impossible in mini-invasive techniques. A tensiometer is inserted and medial and lateral soft tissues are tensioned. Blocks are used to determine femoral component size, and the anterior cut is completed. Figure 26-4 shows the positioning of a tensiometer in combination with
P.185
a cutting block to demonstrate how an equal flexion gap could be created using this technique.
Appropriate posterior cruciate ligament (PCL) balancing is critical to prevent stiffness as well as instability in PCL-retaining implants. As mentioned previously, a tight PCL will limit flexion and may contribute to postoperative stiffness and increased polyethylene wear. Typically, lift-off or rollback is observed. A PCL release using a “pie crust” technique of tight fibers, a superior release of the origin, or posterior release of the insertion easily balances the flexion gap. An overreleased PCL may lead to an incompetent ligament, with paradoxic roll-forward of the femur and flexion instability. Flexion instability after cruciate-retaining TKA has been reproducibly treated with revision to a posterior stabilized TKA with careful balancing of the flexion and extension gaps.60
Flexion instability in the setting of posterior stabilized TKA has been classically identified as dislocation of the cam and post mechanism. However, with contemporary implant designs, the jump distance has been significantly increased and the prevalence of frank dislocation is approximately only 0.15%.60 Excessive posterior slope or overresection of the posterior femoral condyles in a PCL-sacrificing TKA can lead to flexion instability because of the associated increased flexion space. This can allow the tibia to subluxate and produce instability with or without dislocation. There have been reports of tibial post fracture, requiring revision. Post failure may be related to increased stresses on the tibial post in the setting of a loose flexion space.61 Reconstitution of the flexion space in revision TKA for flexion instability is accomplished with the use of posterior femoral augmentation combined with a larger femoral component. Alternatively, more distal femoral bone resection can be made if the bone stock allows, creating symmetric flexion and extension gaps that can be filled with a larger polyethylene insert.60
Patellofemoral Instability
The most common causes of pain and the most commonly cited reasons for revision TKA are complications involving the extensor mechanism and patellofemoral joint. Historically, patellofemoral instability after TKA ranged from 10% to 35%. Recent improvements in prosthetic design and surgical technique have lowered these rates to 1% to 12%.67 Complications include patellar subluxation or dislocation, patellar component wear, and loosening.
Malrotation of the femoral and tibial components is one of the most frequent causes of patellofemoral complications. Limb alignment, preparation of the patella, prosthetic design, and soft tissue balance all contribute to the stability of the patellofemoral joint. Nonsurgical treatments such as bracing and physical therapy are rarely effective in correcting structural abnormalities that lead to patellofemoral maltracking.67 Treatment should be directed by the cause. Computed tomography (CT) scan is the most accurate and reliable way to assess component positioning and its impact on stability. One study using CT to analyze component rotational alignment found that the combined amount of internal rotation of femoral and tibial components correlated directly with the severity of patellofemoral instability.64 If malposition is present, revision of one or both components may be indicated. Lateral retinacular release, with or without vastus medialis advancement may also help align the extensor mechanism axis.
Nerve Injury
Neurologic injury after TKA is an uncommon but potentially devastating complication. Multiple retrospective studies examining large consecutive series (>1,000) identify an incidence of 0.3% to 1.3%.65,68 Subclinical palsy may occur in higher numbers but may be diagnosed only by electrodiagnostic testing.
Predisposing Factors
The cause of peroneal nerve palsy is multifactorial, and no definitive causal relationships have been documented. Conditions associated with peroneal nerve injury include severe flexion and valgus deformity correction, preoperative neuropathy, postoperative epidural analgesia, external leg compression, tourniquet time, and rheumatoid arthritis (RA).
Early studies support the finding that correction of severe valgus and flexion contractures is associated with increased postoperative peroneal nerve palsy. The average preoperative valgus in the patients who developed peroneal palsy ranged from 18 degrees to 23.3 degrees, and average flexion contracture ranged from 15.5 degrees to 22 degrees.65 The incidence ranges from 3% to 10% for correction of knees with severe valgus and flexion deformity. The mechanism of nerve injury has been suggested to relate to narrowing of the extraneural and intraneural microvasculature associated with stretching of the nerve within its surrounding soft tissue.55 An anatomic study examining the risk of direct injury to the nerve during releases to allow for correction of valgus deformity measured a mean bone to nerve distance of 1.49 cm at the level of the standard tibial resection. Those authors concluded that the nerve is adequately protected at the posterolateral corner of the knee, but that care should be taken when performing a “pie crust” release.55
Epidural analgesia postoperatively may be a risk factor. The sensory block may allow the patient to position the leg in a way that directly compresses the nerve. Also, the patients may tolerate excessive motion in extension or overly constrictive dressings leading to nerve palsy. The epidural may mask a peroneal nerve palsy occurring at the time of surgery and delay diagnosis and initiation of treatment.
Prior neuropathy, both central and peripheral, has been associated with development of peroneal palsy. It is theorized that nerves with prior compromise are more susceptible to a second insult, often termed the “double-crush” phenomenon.65 No studies have associated diabetic neuropathy with increased risk for peroneal nerve palsy.
Several studies have identified increased rates of peroneal palsy in patients with rheumatoid arthritis.65,68 In a study by Schinsky et al.,6853% of patients who developed peroneal palsy had a diagnosis of RA, which was significantly higher than the prevalence of RA in their cohort of 1,476 patients. Their patients did not have higher amounts of preoperative valgus or flexion contracture, suggesting that peroneal palsy in RA patients may be via a mechanism unrelated to the deformity of the knee.
P.186
Elevated tourniquet times have been associated with electromyogram (EMG) changes in the peroneal and tibial nerves, but the clinical significance of these changes is unclear. The mechanism is thought to be related to both ischemia and mechanical deformation, but most changes have been identified directly beneath the cuff. One study identified tourniquet time of >120 minutes as an independent risk factor for peroneal palsy.65 Tourniquet release, allowing a reperfusion interval of 10 to 30 minutes, followed by reinflation has been recommended to extend the duration of tourniquet time if needed.69 One recent study reviewed a consecutive series of >1,000 patients undergoing TKA with a tourniquet time of >120 minutes. The overall incidence of neurologic complications was 7.7% in this population. Complete neurologic recovery occurred in 89% of patients with peroneal nerve palsy.69
The peroneal nerve is vulnerable to direct compression, given its superficial anatomic location as it winds around the fibular head. Direct compression on the peroneal nerve by constrictive dressings has been suggested to play a role in the development of palsy. In addition, the development of postoperative hematoma has been identified as a rare source of compression on the nerve leading to palsy.
Treatment and Prognosis
Standard nonoperative management of peroneal nerve palsy includes immediate removal of any constrictive dressings and flexion of the hip and knee to approximately 20 degrees and 45 degrees, respectively. This can be accomplished by elevation of the extremity on several pillows. Initial short-term treatment should be observation and an ankle/foot orthotic (AFO) device for foot drop. Surgical exploration has been indicated if no functional recovery is noted after 3 months from onset, particularly if the AFO is not tolerated. The routine use of surgical decompression remains controversial, despite several reports of full recovery following open exploration.
The potential for recovery after peroneal nerve palsy following TKA ranges from 50% to 89%.65,69 The less severe the initial palsy, the more likely it is to completely resolve. Despite varying percentages of complete peroneal nerve recovery, most patients have demonstrated good functional capacity after TKA.
References
P.187
Suggested Readings
Chauhan SK, Scott RG, Breidahl W, et al. Computer-assisted knee arthroplasty versus a conventional jig-based technique. A randomised, prospective trial. J Bone Joint Surg Br. 2004;86: 372–377.
Haas SB, Cook S, Beksac B. Minimally invasive total knee replacement through a mini midvastus approach: a comparative study. Clin Orthop Relat Res. 2004;428:68–73.
Laskin RS, Beksac B, Phongjunakorn A, et al. Minimally invasive total knee replacement through a mini-midvastus incision: an outcome study. Clin Orthop Relat Res. 2004;428:74–81.
Berend ME, Ritter MA, Meding JB, et al. Tibial component failure mechanisms in total knee arthroplasty. Clin Orthop Relat Res.2004;428:26–34.
Sparmann M, Wolke B, Czupalla H, et al. Positioning of total knee arthroplasty with and without navigation support. A prospective, randomised study. J Bone Joint Surg Br. 2003;85:830–835.
Stevens CM, Tetsworth KD, Calhoun JH, et al. An articulated antibiotic spacer used for infected total knee arthroplasty: a comparative in vitro elution study of simplex and palaces bone cements. J Orthop Res. 2005;23(1):27–33.
Wasielewski RC, Galat DD, Komistek RD. Correlation of compartment pressure data from an intraoperative sensing device with postoperative fluoroscopic kinematic results in TKA patients. J Biomech. 2005;38(2):333–339.