Thromboembolism

DVT occurs in approximately 50% of unilateral cases and in 75% of bilateral cases when no prophylaxis is used. Although DVT occurs mainly in the calf veins (Fig. 125-1), life-threatening emboli do not arise from this region. In contrast to the situation after total hip arthroplasty (THA), isolated proximal vein thrombosis does not seem to occur after knee surgery, despite the possible trauma of a pneumatic tourniquet.

Figure 125-1 Venogram showing clots in the calf veins.

Some have argued that distal thrombosis can be ignored, provided that the patient convalesces normally and is not confined to bed for a lengthy period. We believe that this view is too optimistic, although we concede that the risk of a fatal thromboembolus is small and is lower than that after THA. In a prospective review of 527 TKAs in 499 patients, Khaw and colleagues,¹¹⁸ using no prophylaxis other than antithrombotic stockings and relatively early (48 hours) mobilization, found only one death in 3 months from pulmonary embolism (0.19%). This patient had bilateral TKA and had a myocardial infarction 1 day postoperatively, with the subsequent pulmonary embolus occurring 22 days postoperatively. Seven other patients (1.3%) developed symptomatic pulmonary embolism and were treated with anticoagulation, without sequelae. This low incidence of fatal pulmonary embolism has been supported in other studies by Stulberg and associates (0%),²¹⁴ Stringer and coworkers (0%),²¹³ Khaw and colleagues (0.2%),¹¹⁸ and Ansari and associates (0.4%).⁴ Fatal emboli do occur, however, and three patients died as a result of emboli in the first 400 arthroplasties performed at the Hospital for Special Surgery. No specific prophylaxis was used at that time; the cause of death was confirmed by autopsy in all cases. In addition, we believe that the incidence of fatal pulmonary emboli is underestimated. A certain proportion of calf clots propagate proximally to form the more dangerous clots in the popliteal and femoral veins. This process takes time, however, and the major risk period for a fatal pulmonary embolism after TKA may be in weeks 3 and 4 postsurgery; this is in contrast to what occurs after THA, in which a clot in the proximal veins may be found in 20% of cases after the first week. Because most TKA patients now are discharged after hospitalization of 2 to 3 days, sudden death at home in a patient having no clinical evidence of vein thrombosis may be attributed to myocardial infarction.

Calf clots may not in themselves be important, but should be regarded as a harbinger of more proximal clotting. Haas and coworkers⁸⁵ have studied 1329 patients with 1697 TKAs. Thrombosis was found in 808 patients (61%); 53% had thrombosis of the calf vein and 8% had thrombosis of the proximal veins. The lung scans of 60 patients (4.5%) were positive, and symptomatic pulmonary emboli occurred in 14 patients (1.1%). All these patients received aspirin as prophylaxis. Venography was performed between postoperative days 4 and 6. A perfusion lung scan obtained on postoperative day 5 to 7 was compared with a preoperative baseline perfusion lung scan. Thrombosis of the calf vein was treated with warfarin (Coumadin); the dosage was adjusted to maintain a prothrombin time at approximately 1.5 times control. Only patients with symptomatic proximal thrombi or symptomatic pulmonary emboli were fully anticoagulated with intravenous heparin. Although the natural history of thromboembolic disease was altered by prophylaxis and treatment, 6.5% of patients with calf thrombi had a positive lung scan and 1.6% had symptomatic pulmonary emboli. This finding was compared with patients who had no venographic evidence of deep vein thrombosis (DVT), of whom 1.9% had a positive lung scan and 0.2% had symptomatic pulmonary emboli. Statistical significances for the difference in lung scan and pulmonary embolus results were P = .001 and P = .034, respectively. Patients with calf or proximal thrombi were found to have similar rates of positive lung scans and symptomatic pulmonary emboli.

The risk of symptomatic and fatal pulmonary embolism is of greatest concern. Without prophylaxis, the rates of pulmonary embolism and fatal pulmonary embolism range from 1% to 28% and 0.1% to 2.0%, respectively.⁷⁴ From six randomized studies comparing low-molecular-weight heparin (LMWH) with warfarin, the total rate of DVT was 33% for patients who received LMWH compared with 48% for patients who received warfarin. Similarly, the rate of proximal DVT was 7.1% and 10.4% for patients who received LMWH and warfarin, respectively. Barrett and colleagues¹¹ have reviewed 122,385 U.S. Medicare enrollees who had a TKA in 2000. Pulmonary embolism developed in the first 3 months in 0.81% of patients who had a unilateral TKA compared with 1.44% of patients who had undergone a simultaneous bilateral TKA.

The discussion of appropriate DVT prophylaxis is covered elsewhere in this text and is beyond the scope of this chapter. Many questions concerning how best to effect thromboprophylaxis in the orthopedic setting remain unanswered, and our (likely unobtainable) need for an ideal strategy combining complete efficacy with absolute safety remains unmet. However, we should not permit such questions and needs from obscuring fundamental truths made clear by intense research over the past several decades: (1) venous thromboembolism (VTE) is a known and serious complication of total joint arthroplasty; (2) evidence-based thromboprophylaxis works; (3) the orthopedic community needs to improve compliance with available guidelines and work on reaching a consensus.

Fat Embolism Syndrome

The diagnosis of fat embolism can be elusive,^{20,32,56,64,160} and we suspect that the condition often may pass unrecognized as the cause of transient confusional states after surgery. The syndrome results from the embolization of fat and other debris from the femur or tibia that travels mostly to the lungs. The initial effects in the lungs are mechanical, with an increase in perfusion pressure, engorgement of the vessels in the lungs, and secondary right-sided heart strain. In the presence of hypovolemic shock, the patient may die from acute right-sided heart failure. The delayed effects of fat embolism occur after 48 to 72 hours because of the chemical effects of fat. The pulmonary tissue secretes lipase, which hydrolyzes fat into free fatty acids and glycerol. These free fatty acids increase capillary permeability, cause destruction of alveolar architecture, and damage lung surfactant. The end result of all these changes is hypoxia.¹⁸³

Clinical findings of fat embolism syndrome include tachypnea, dyspnea, profuse tracheobronchial secretions, apprehension, anxiety, delirium, confusion, unconsciousness, and petechial hemorrhage. Laboratory findings are hypoxemia on arterial blood gas testing and thrombocytopenia lower than 150,000 mm³. Treatment is supportive. Mechanical ventilation may be required in advanced cases. Corticosteroids may be beneficial in diminishing the inflammatory response from the chemical effects of fat emboli.¹⁷⁴

Monto and coworkers,¹⁶⁰ in a review of the literature, reported 19 cases with 9 deaths. Of the 19 cases, 15 were associated with long-stem cemented prostheses, such as the Guepar hinge. Four cases were associated with total condylar arthroplasty. One case was associated with intramedullary instrumentation. Fahmy and colleagues⁶⁴ have shown human intramedullary (IM) femoral canal pressures of 500 to 1000 mm Hg generated by using standard alignment rod techniques. Venting the canal did not lower the canal pressure significantly. IM pressures were maintained within normal limits only by overdrilling the femoral canal and gently placing the guide rod. Copious irrigation of the IM canal with pulsatile lavage, suction of marrow contents after irrigation, and use of fluted rods to assist the egress of bone marrow elements may reduce the intramedullary pressure. The use of a pneumatic tourniquet does not protect against fat embolism and, with the popularity of IM guidance systems for femoral and tibial components, an increase in the incidence of fat embolism syndrome may be expected. The surgeon should consider avoiding routine instrumentation of the IM canals of the femur and tibia in bilateral cases.

Young-Hood and Kim and colleagues¹²¹ have compared the prevalence of fat embolism after TKA in 160 patients (210 knees) with navigation and 160 patients (210 knees) without navigation. They found no significant difference in the intraoperative and postoperative hemodynamic values between the groups. A higher prevalence of fat embolism was seen (60% vs. 41%) in patients with higher triglyceride levels. However, in contrast, Kalairajah and colleagues¹¹³ have demonstrated a significant reduction in the number of cranial fat emboli in a group treated with computer-assisted TKA compared with a group treated with TKA using standard IM instrumentation.

Wound Drainage and Delayed Wound Healing

Soft tissue considerations must always be at the forefront when planning surgical intervention. This is especially true for high-risk patients, including those with prior incisions. Although several plastic surgery techniques are available to treat wound complications, much of the damage is already done once the complication occurs. Even the best outcomes of salvage techniques following wound failure result in functional loss and cosmetic deficit. Therefore, the surgeon’s primary goal should be to avoid postoperative wound complications. Despite appropriate planning and meticulous technique, however, wound complications will occur. The orthopedic surgeon must work in conjunction with the plastic surgeon and be aware of the nonsurgical and surgical techniques available to minimize functional loss. Inappropriate management of postoperative skin problems can result in failure of the reconstruction, deep infection, a nonfunctioning extremity, amputation, and/or a potentially life-threatening situation.

Postoperative Considerations

Appropriate postoperative wound management depends on the severity and timing of the complication. The failure of the soft tissues to heal following TKA has serious effects. Careful examination at the time of the first dressing change can often alert the clinician to a potentially problematic wound. Early indications may include ecchymosis, blistering, and persistent or large amounts of wound drainage. The goal is early intervention, when possible, to prevent further wound breakdown and complication.

Serosanguineous drainage from the incision is common and is a cause for concern only when the drainage is profuse and persistent. If there is no purulence or erythema, initial management should be a compression dressing, immobilization, and observation. Wound healing takes precedence over motion. When there is drainage, antibiotic therapy is controversial because it may mask a deep infection, making it difficult to identify the causative organism. We recommend consulting with an infectious disease specialist and administering intravenous antibiotics only in select cases. This is a matter of clinical judgment, but we do not consider antibiotics in these circumstances wrong if clinical suspicion for a deep infection is low. Prolonged and persistent serosanguineous drainage raises the issue of a capsular defect, which should be surgically repaired. Some authors recommend that if drainage does not stop after 5 days of appropriate treatment, an open débridement should be performed.

Local Care

Local care measures begin with frequent dressing changes, elevation of the extremity, and limiting mobilization, including ROM activities. When a wound is identified to be at risk in the early postoperative period, we routinely discontinue continuous passive motion, apply a compressive dressing, place the knee in an immobilizer, and institute physical therapy. When superficial epidermal loss occurs in an area smaller than 2 to 3 cm², several modified dressing protocols can be instituted to protect the underlying tissues and allow for secondary healing (Fig. 125-2). These include antibacterial ointments or gels and enzymatic débriding agents. During this phase, we may consider temporary discontinuation of the anticoagulant agent, because a hematoma related to overaggressive anticoagulation can be devastating at this stage of healing. The use of mechanical devices for DVT prophylaxis should be considered until the wound stabilizes.

Figure 125-2 Because of its subcutaneous position, skin necrosis is a particular hazard in knee joint replacement. Overaggressive débridement can lead to deep infection. When necrosis occurs, the knee should be immobilized until the eschar separates.

Irrigation and Débridement

Early surgical intervention may be helpful and is indicated in certain situations. In cases of imminent wound compromise caused by a large or expanding hematoma or prolonged wound drainage beyond 1 week postoperatively, early surgical interventions should be considered. The goal of irrigation and débridement (I&D) is to prevent further wound breakdown and deep infection. Studies have shown that each day of persistent drainage greatly increases the risk of wound infection. In addition, other studies have shown a lower incidence of deep infection when postoperative hematoma or persistent drainage is treated with I&D versus nonsurgical management.

Burnett and coworkers³⁰ have reported that the rates of surgical site complications necessitating readmission, irrigation and débridement of hematoma and the wound, or prolonged hospitalization for wound drainage are 4.7%, 3.4%, and 5.1%, respectively. Wound drainage occurred for 4 to 7 days after 9.3% of the procedures and for more than 7 days after 9.3% of the procedures, with more than 7 days of drainage being highly predictive of readmission and wound reoperation.

Once the decision is made to proceed with surgery, the patient is taken to the operating room expediently. The setup is the same as with the index arthroplasty. Thorough I&D is performed. Deep cultures should be obtained to direct antibiotic therapy, if indicated. If an opening in the arthrotomy exists, or if the hematoma is deep to the arthrotomy, then the entire prosthesis should be exposed and irrigated and a polyethylene liner exchange is performed. It is best to assume that deep infection is present and perform a thorough I&D with antibiotic irrigation. After hemostasis is obtained, the wound is closed in a layered fashion over a drain. A compressive dressing is applied. Occasionally, when a tension-free closure cannot be obtained, a gastrocnemius muscle flap may be necessary. This should be anticipated before surgery. Appropriate accommodations should be made preoperatively, including the involvement of a plastic surgeon.

Postoperatively, the limb is elevated while ROM and chemical DVT prophylaxis are initially held. Broad-spectrum antibiotics are given until pending cultures are final. An infectious disease consultation is obtained for patients in whom infection is suspected or cultures return positive. Decisions regarding range of motion, DVT prophylaxis, and continued antibiotic therapy are made on an individual basis.

Galat and colleagues,⁷² in their study of 17,784 primary TKAs, have reported that the rate of early return to surgery for wound complication is 0.33%. They found that the 2-year cumulative probabilities of major subsequent surgery (component resection, muscle flap coverage, or amputation) and deep infection were 5.3% and 6.0%, respectively, in the knees with early surgical treatment of wound complications as compared with 0.6% and 0.8%, respectively, for knees without early surgical intervention. The overall prosthesis salvage rate was 98%, but obviously results are inferior compared with those of patients without early wound complication.

Hematoma

A current debate exists between appropriate VTE guidelines following TKA. Although discussed elsewhere in this text, the debate centers around the incidents of postoperative hematoma. Given the nature of a TKA with numerous bone cuts, intramedullary canal violation, and soft tissue releases, it is no surprise that postoperative hematomas can occur. Symptoms include not only intense palpable hemarthrosis around the knee but can also be accompanied by skin discoloration, bruising, increased pain, decreased ROM, and wound drainage.

Wound drainage is of particular concern. Weiss and Krackow²²⁸ have described an increased rate of knee infection with prolonged wound drainage. Patel and associates¹⁷³ have evaluated factors that lead to prolonged drainage in the postoperative period. The authors have demonstrated that an increase in the initial postoperative drainage in the recovery room was most indicative for further drainage on days 2, 3, 4, and 5. In other words, better control in the immediate postoperative period, with less postoperative recovery room drainage, led to less wound drainage on subsequent hospitalization days. Limiting blood loss in the recovery room begins in the operating room (OR). Plugging of the intramedullary canal can help decrease initial blood loss as can cauterization of known sources of bleeding, such as genicular vessels and the posterior capsule⁴⁹ (Fig. 125-3). Intraoperative tourniquet deflation has not been shown to decrease blood loss; in fact, some studies have shown an increase in blood loss when this is performed.¹⁸² The appropriate timing of tourniquet release remains debated, even among surgeons at our institution. The improved designs of TKA and improved surgical techniques have led to improved patellar tracking, resulting in less wound drainage when the lateral release is avoided. At our institution, we use a multimodal approach to minimizing intraoperative bleeding, including a smaller quadriceps-sparing incision, meticulous surgical technique, lidocaine and epinephrine injection, bipolar sealer device, and a topical fibrin spray.

Figure 125-3 Cauterization of the lateral genicular artery. Unless this artery is visible, it is advisable to release the tourniquet before closure so that the vessel can be clearly identified. It is the major source of dangerous postoperative bleeding.

When postoperative hematoma occurs, a period of immobilization and observation is permissible, and many hematomas subside spontaneously. If bleeding continues, as evidenced by tense and painful swelling, the knee must be reopened and the source of the bleeding identified (an argument for intraoperative tourniquet release). Probing and squeezing the wound to evacuate a hematoma are not recommended because this could lead to retrograde contamination. Indications for surgical evacuation of a hematoma include skin compromise, wound dehiscence, impending skin necrosis, leakage through the incision, persistent sanguineous drainage, and pain. Galat and associates⁷¹ have studied 42 TKA patients matched with 42 control subjects. They found that the rate of return to surgery within 30 days for evacuation of a postoperative hematoma was 0.24%. The 2-year rates of a subsequent major operation or deep periprosthetic infection in the study group were 12.3% and 13.6%, respectively and 0.9% and 1.4%, respectively, in the control group.

Skin Necrosis

Postoperative soft tissue and skin necrosis should be handled early and aggressively. Treatment options for Small areas (<4 cm), especially over the patella or more proximal, include healing by secondary intention, split-thickness skin grafts, or local fasciocutaneous flaps. Small areas (<4 cm), especially over the patella or proximal, may be treated by secondary intention healing, split-thickness skin grafts, or local fasciocutaneous flaps. Skin loss over the patellar tendon or tibial tubercle is best treated with muscle flap coverage; usually, the medial head of the gastrocnemius rotator flap is adequate.

Anatomy of the Peroneal Nerve

The peroneal nerve is composed of fibers from the dorsal portion of the L4 and L5 and S1 and S2 nerves. As the nerve courses down from the thigh, it curves laterally behind the head of the fibula to reach the two heads of the peroneus longus. The nerve flattens as it passes between these two heads, separating the bundles and exposing unprotected nutrient vessels. The nerve then curves around the neck of the fibula and divides into deep and superficial branches. The deep peroneal nerve continues under the extensor digitorum longus, along the anterior aspect of the intraosseous membrane. It sends motor branches to the tibialis anterior, extensor digitorum longus, extensor hallucis longus, and peroneus tertius. The nerve continues distally, ending in the medial and lateral terminal branches, which, among other functions, supply sensation to the first web space of the foot.

The superficial peroneal nerve passes distally between the peronei and extensor digitorum longus. Motor branches are given off to the peroneus longus and peroneus brevis. In the lower third of the leg, the nerve branches into the medial and intermediate dorsal cutaneous nerves. These terminal branches complete the sensory innervation of the feet. Bruzzone and associates²⁶ identified anatomic landmarks after 20 cadaveric knee dissections and defined a danger zone and a safe zone to avoid common peroneal nerve injury when performing the inside-out release technique of the posterior-lateral corner during TKA. During the release of the posterior-lateral capsule, the nerve was identified to be at risk in the triangle defined by the popliteus tendon, tibial cut surface, and most posterior fibers of the iliotibial band (danger zone), but not during “pie-crusting” of the iliotibial band (safe zone). They also defined the average distance from the nerve to the posterior-lateral corner of the tibia and to the posterior border of the iliotibial band as 13.5 and 35.8 mm, respectively.

Postoperative Clinical Findings

Peroneal nerve palsy is an infrequent but worrisome complication after TKA. The incidence is higher in revision cases than primary TKA. The prevalence of peroneal nerve palsy has been reported many times in the literature. Mont and coworkers¹⁵⁷ have reviewed the literature and found the cumulative prevalence to be 0.58% (74 of 12,784). Asp and Rand⁶ reported an incidence of 0.3% in 8754 TKAs performed at the Mayo Clinic. Yacub and colleagues,²⁴⁰ in their retrospective study of 14,979 patients, reported a 0.01% incidence of nerve injury after TKA. They noted a 10-fold difference in nerve injury rates between diabetic and nondiabetic patients, 0.11% versus 0.01%, respectively.

Numerous causes for nerve palsy have been described, but they are most commonly associated with the correction of severe flexion and valgus deformity or a combination of the two. The following factors contribute to the development of peroneal nerve palsy: (1) stretching of the nerve in valgus and flexion contractures; (2) fascial compression of the nerve and its vascular supply; (3) direct pressure from the postoperative dressing; and (4) epidural anesthesia. In rare idiopathic cases, none of the mechanisms listed seem to apply. Factors that have not been found to be associated with peroneal nerve palsy include age, gender, type of arthritis, and duration of tourniquet. Idusuyi and Morrey¹⁰³ performed 10,321 TKAs from 1979 through 1992 at the Mayo Clinic and reported 32 postoperative peroneal nerve palsies. The factors associated with peroneal nerve palsies in this series were epidural anesthesia for postoperative pain control, previous laminectomy, and preoperative valgus deformity. The relative risk for patients who had previous proximal tibial osteotomy was doubled but was not statistically significant.

In patients with severe combined deformities, the large soft tissue dissections required to balance the deformities may be responsible for increased traction or vascular compromise to the nerve. The alternative method of bone sacrifice to correct large deformities, although appealing because of elimination of the need for obtaining soft tissue balance, does not entirely avoid peroneal nerve palsy. The bone-sacrificing method also leaves a residual, permanent, leg-length discrepancy that may be associated with an extensor lag because of relative lengthening of the quadriceps mechanism. When ligament balance is not attempted, a constrained prosthesis is needed. In younger patients, this may lead to eventual loosening, although, with the constrained condylar prosthesis, this has not yet been seen in our practice. In older patients at high risk for peroneal nerve palsy, the use of the constrained condylar knee is a suitable alternative. These patients recover much more rapidly than after major release procedures and, because of their age, the risk of ultimate loosening is less concerning.

Although it is sometimes necessary to splint the knee in extension after correction of a severe flexion contracture, evidence of nerve palsy demands immediate removal of the splints and flexion of the knee. If this is done promptly, complete recovery can be expected. In the case of varus deformities, splinting is not necessary. For severe valgus deformity (>20 degrees), one may consider splinting the patient with the knee in 30 degrees of flexion. Nerve function is then monitored closely when the surgical dressing is removed on postoperative day 2. Tight dressings that might press directly on the peroneal nerve should also be avoided.

Postoperative dressings that are too tight or improperly padded can cause direct pressure on the peroneal nerve, causing nerve palsy. Beller and associates,¹⁵ in their study, described a double crush syndrome, which includes unrecognized pressure on the peroneal nerve caused by continuous epidural anesthesia and an axonal lesion from the pressure of the pneumatic tourniquet. To prevent a peroneal lesion after TKA while using continuous epidural anaesthesia, they recommended limiting the pneumatic tourniquet pressure to 320 mm Hg and ensuring pressure-free positioning of the operated leg.

Asp and Rand⁶ performed 8998 arthroplasties and reported 26 nerve palsies. They noted that complete recovery was more likely when the palsy was initially incomplete. Dressing removal and flexion of the knee on diagnosis did not always help. The time of presentation of the problem varied from discovery in the recovery room to postoperative day 6. The motor fibers of the tibialis anterior and extensor hallucis longus muscles were affected in all cases; a sensory deficit was noticed in 20 patients (87%). The peroneus longus muscle was affected in nine patients (39%). Electromyographic evaluation showed a diffuse motor neuropathy in the mild cases and denervation potentials in muscles supplied by the deep branch of the common peroneal nerve in the more severely involved cases. The treatment rendered on discovery of these findings varied according to the discretion of the surgeon involved. The most frequent therapeutic measure was to loosen the Robert Jones dressing and place the knee in a more flexed position. In two cases, this maneuver brought immediate improvement of motor and sensory deficits. The time interval between discovery and the beginning of the return of function ranged from immediately in the two cases mentioned to 6 months. Motor improvement occurred first, with sensory return lagging behind.

Omeroglu and coworkers¹⁶⁶ have reported the case of 46-year-old rheumatoid patient with bilateral peroneal nerve palsy on the second postoperative day after simultaneous bilateral TKA. Electromyographic studies revealed bilateral axonotmesis. Complete motor recovery was seen on both sides within 6 months, although sensory deficit was present on one side at 2 years postoperatively. They reported preoperative severe flexion contracture and epidural anesthesia as the risk factors for the development of the nerve palsy in this patient.

Treatment and Results

The treatment of chronic peroneal nerve palsy usually consists of an ankle-foot orthosis for a footdrop and passive ankle ROM to prevent an equinus deformity. Complete recovery of a peroneal palsy is rare. The recovery seen is usually partial, with sensory deficits that may be permanent; residual motor deficits are not usually of clinical significance. Occasional marked weakness, especially of the great toe extensor, may be seen. In Asp and Rand’s study⁶ of 26 postoperative peroneal palsies, palsies were complete in 18 and incomplete in 8. Of these patients, 23 had motor and sensory deficits and 3 had only motor deficits. At 5-year follow-up, recovery was complete for 13 palsies and partial for 12. Complete recovery was more likely in palsies that were initially incomplete.

Although most investigators support nonoperative treatment of this complication, others disagree. Krackow and colleagues¹²⁹ treated five patients with operative exploration and decompression of the peroneal nerve for a postoperative palsy. The procedure was performed 5 to 45 months after the index TKA. All patients have shown improved nerve function, and 4 of 5 patients had full peroneal nerve recovery. All patients were able to discontinue their ankle-foot orthosis.

Instability

Instability is a common cause of mechanical failure of TKA. Instability accounts for 10% to 22% of TKA failures requiring revision.* The direction of instability at the tibiofemoral articulation can occur in the coronal (varus-valgus) plane, sagittal (anteroposterior) plane, or as a combination of planes. Early instability may be a result of malalignment of the components, failure to restore the mechanical axis of the limb, imbalance of the flexion-extension space, intraoperative or postoperative rupture of the medial collateral ligament (MCL), or posterior PCL rupture with cruciate-retaining designs. Commonly, late instability is secondary to polyethylene wear. Asymmetrical polyethylene wear related to malalignment can result in relative lengthening of the collateral ligament on the involved side and subsequent coronal instability. In patients with cruciate-retaining knees, it is not uncommon for the PCL to elongate or attenuate. This can lead to progressive polyethylene wear and late sagittal plane instability. However, late sagittal plane instability is seen not only with cruciate-retaining designs, because significant wear or fracture of the tibial polyethylene post in posterior-stabilized knees may result in late sagittal plane instability.²⁰⁷ Instability after knee arthroplasty can be subdivided into three types—extension instability, flexion instability, and genu recurvatum.

Extension Instability

Symmetrical Instability

Symmetrical extension instability occurs when the extension space is not filled by the thickness of the components. This can be caused by overresection of the tibia or the distal femur.²⁴ Although overresection of the tibia results in equally large flexion and extensions gaps, excessive resection from the distal femur results in asymmetrical flexion and extension gaps. If overresection of the tibia is recognized intraoperatively, it can be easily managed by increasing the size of the polyethylene insert. Similarly, if a postoperative TKA is noted to be equally loose in flexion and extension, one option is revision surgery with a polyethylene exchange to a larger insert. Overresection of the distal femur creates a much different scenario. With overresection of the distal femur, the joint line is elevated. This creates a larger extension gap compared with the flexion gap. By simply increasing the size of the tibial insert, the result will be a knee that is too tight in flexion. The elevated joint line will also affect patellar tracking and limit flexion, and may result in midflexion instability¹⁷⁰ (Fig. 125-5). It is best in these cases to restore the joint line by adding distal femoral augmentation.

Figure 125-5 If an extra thick tibial component is needed to stabilize the knee, the patella is displaced distally, causing a patella infra. Undersizing of the femoral component in the sagittal plane and anterior malpositioning are possible causes. Excessive distal resection of the femur moving the joint line proximally is another cause.

Asymmetrical Instability

Asymmetrical extension instability is frequently encountered. Following the femoral and tibial preparation, the extension and flexion gaps must be assessed. An asymmetrical extension gap in the face of appropriate bony resection is most often caused by inadequate correction of the preoperative deformity.^24,170 Preoperative varus deformity is the most common soft tissue deformity. The shortening or tightening of the medial structures, including the superficial MCL, must be recognized. Often, these tightened structures are not sufficiently released and the deformity is not completely corrected (Fig. 125-6). This will lead to progressing varus deformity with increased medial joint line stresses, accelerated polyethylene wear, and stretching out of the soft tissue structures on the lateral side. The asymmetrical extension gap must be recognized intraoperatively and an appropriate subperiosteal elevated of the superficial MCL performed, as described by Insall and associates.¹⁰⁵ They thought that valgus instability from overrelease of the medial structures during correction of a fixed varus deformity was rare.

Figure 125-6 A, This arthroplasty is unstable, but not because of a ligament deficiency. B, The medial ligament was not released; an asymmetrical instability remains.

Knees with asymmetrical extension instability may also result from undercorrection of a preoperative valgus deformity.¹⁷⁰ Excessively tight lateral structures, including the lateral collateral ligament (LCL) and iliotibial band (ITB), if left uncorrected, will lead to recurrence of the valgus deformity and increased lateral joint line forces. Currently, the most commonly used technique to correct preoperative valgus deformity without causing lateral instability has been described by Insall and associates¹⁰⁵ as the “pie-crust” technique. This is recommended for patients with a preoperative valgus deformity of less than 20 degrees.³ Clarke and coworkers³⁶ have published results of 24 TKAs with preoperative valgus deformity ranging from 9 to 30 degrees. The deformity was corrected using the “pie-crust” technique and the knee was reconstructed with a posterior-stabilized prosthesis. This series reported complete correction of the deformity and no instability in all cases. In patients with more than 20 degrees of valgus deformity, the risks of iatrogenic peroneal nerve injury must be considered.⁵⁹ The surgeon may consider the use of a constrained condylar design in much older patients, whereas increased constraint in younger patients should be avoided unless absolutely necessary.

Extension instability can result secondary to iatrogenic injury of the collateral ligaments (Fig. 125-7). The excursion of large oscillating saw blades was studied and found to be greater than the width of the medial and lateral femoral condyles in the female knee.⁵⁹ Meticulous attention must be paid to the position of retractors during femoral and tibial resection. Also, iatrogenic collateral ligament injury can be produced during aggressive testing of varus-valgus stability or with attempts to reduce the polyethylene insert in an excessively tight knee. With iatrogenic injury to the MCL, most authors advocate a direct repair with Krackow-pattern suturing. If primary repair is insufficient, augmentation can be performed using the hamstring tendons. Finally, a constrained condylar design can be used to provide additional stability while the repair heals. Several studies have demonstrated good outcomes without the use of increased constraint or tissue augmentation. Leopold and coworkers¹³⁷ have demonstrated good to excellent results in 100% of 16 TKAs treated with direct repair of an iatrogenically injured MCL, reconstruction using a cruciate-retaining implant, and postoperative bracing for 6 weeks. More recently, Koo and Choi¹²⁶ performed a retrospective review of 15 TKAs treated conservatively for intraoperative iatrogenic detachment of the MCL from the tibia. Specifically, no increased constraint, bracing, or ligament reconstruction was performed. No significant difference was seen between MCL-intact TKAs and the knees with iatrogenic MCL injury with a minimum of 2-year follow-up.

Figure 125-7 Radiograph showing the instability that results from a transected medial collateral ligament. This is not the result of an overzealous medial release.

Flexion Instability

Flexion instability results from inadequate filling of the flexion gap or attenuation of the PCL following cruciate-retaining TKA. Flexion instability is now a well-recognized cause of TKA failure (Fig. 125-8). Early flexion instability is likely secondary to gap imbalance, with or without laxity of a collateral ligament¹⁷⁰ (Fig. 125-9). Late flexion instability can occur in cruciate-retaining and posterior-stabilized TKAs, but the cause is different. In patients with cruciate-retaining knees, it is not uncommon for the PCL to elongate or attenuate. This can lead to progressive polyethylene wear and late flexion instability.¹⁶⁷ Late flexion instability in posterior-stabilized knees may result from significant wear or fracture of the tibial polyethylene post.²⁰⁷ Symptoms can range from a vague sense of instability and recurrent effusions to frank dislocation. Flexion instability is best assessed with the knee in 90 degrees of flexion while the patient sits on the end of the examination table.

Figure 125-8 AP radiograph of a knee with flexion instability. With the knee in flexion, a posterolateral dislocation is observed. This dislocation is caused by inequality between the flexion and extension gaps, and the PCL is not sufficient to provide stability, particularly when the prosthesis is unconstrained.

Figure 125-9 In cases with considerable ligamentous laxity, an underresection of the distal femur may be preferable. The standard femoral resection necessitates a thicker tibial component to take up the slack in the soft tissues and may cause distal migration of the patella. By underresecting the femur, a desirable patellar position is maintained.

Frank dislocation following posterior-stabilized TKA is a rare complication²⁰⁴ (Fig. 125-10). Improved prosthetic design with increased jump distances have reduced this complication to well below 0.5%.¹⁷⁰ Although the posterior-stabilized prosthesis can resist direct posterior translation, deep flexion combined with a varus or valgus stress can lead to posterior dislocation of the tibia. Patients who are at highest risk are those who have had the correction of a large valgus deformity and quickly regained their ROM during the postoperative period. The dislocated TKA can continue to function well following closed reduction under anesthesia; however, subsequent dislocation may occur. Recurrent dislocation requires revision to a larger polyethylene insert, if the extension gap permits, or component revision to a constrained condylar design.

Figure 125-10 A, Posterior dislocation of a PS prosthesis. B, After reduction.

More commonly, posterior stabilized knees may be symptomatic because of flexion instability but not demonstrate dislocation.²⁰⁴ Patients report a sense of instability, recurrent effusion, periarticular tenderness, and difficulty ambulating on stairs. The diagnosis is made on clinical examination with the knee flexed at 90 degrees. Schwab and colleagues²⁰⁴ have reviewed 1370 revision TKAs performed at the Mayo Clinic (Rochester, Minn). Ten revisions were performed for flexion instability. Eight (80%) had successful outcomes following revision of both components and appropriate balancing of the flexion and extension gaps. The excessively large flexion gap was frequently managed by upsizing the femoral component following posterior femoral augmentation.

Similarly, cruciate-retaining (CR) knees can be symptomatic because of flexion instability without dislocation. Patients again report a sense of instability, recurrent effusion, periarticular tenderness, pes anserine tenderness, and difficulty ambulating on stairs. Additionally, on physical examination, a posterior sag and positive posterior drawer may be readily apparent. Pagnano and associates¹⁶⁷ have reported two types of causes for flexion instability following CR TKA. The first type creates an excessive flexion gap by surgical error. This can occur by undersizing the femoral component, which results in overresection of the posterior femoral condyles and a reduction in the femoral offset. Alternatively, the creation of excessive tibial slope can result in a knee that is well balanced in extension but remains loose in flexion. The second type involves late failure of the PCL and late instability. Other causes may include polyethylene wear, synovitis, iatrogenic PCL injury, traumatic PCL rupture, and an excessively tight flexion gap and PCL rupture following aggressive ROM exercises.

Treatment of CR knees with flexion instability most often involves revision arthroplasty to a posterior-stabilized (PS) design.¹⁷⁰ At the time of revision, a larger femoral component is frequently used following posterior femoral augmentation. This improves posterior femoral offset and reduces the size of the flexion gap. Attention must also be paid to the tibial slope. If excessive posterior slope is present preoperatively, this must be addressed with revision of the tibial resection. Careful balancing of the flexion and extension gaps is then performed. Pagnano and associates¹⁶⁷ have demonstrated successful outcomes following revision of a CR knee with flexion instability to a well-balanced PS design in 19 of 22 knees (86%).

Related posts:

Management of Extra-articular Deformity in Total Knee Arthroplasty With Navigation Patellar Instability Posterior Cruciate Ligament Reconstruction: Posterior Inlay Technique Anterior Cruciate Ligament Reconstruction With Central Quadriceps Free Tendon Graft Osteochondritis Dissecans Anatomic Aberrations

Stay updated, free articles. Join our Telegram channel

Join