Total knee arthroplasty (TKA) has been shown to be an effective and reliable treatment for end-stage osteoarthritis of the knee. This factor, combined with an aging population, has resulted in an increasing incidence of this procedure, which is expected to continue to rise over the coming decades. , The goals of TKA are to restore patient function and improve pain secondary to degenerative joint disease (DJD), which often occurs later in life and can present synchronously with decreased bone mineral density, osteopenia, and osteoporosis. Given these factors, it is expected that the number of periprosthetic knee fractures will continue to rise over the coming years. These injuries present challenging treatment decisions for the orthopaedic surgeon, as these patients often present with limited bone stock and risk factors for poor fracture healing, such as malnutrition. Nevertheless, the goal of treatment for these patients is to restore preinjury function and maintain a secure, well-balanced, and aligned knee prosthesis.
Types of Periprosthetic Knee Fractures
Periprosthetic knee fractures can be divided according to temporal and location factors. Fractures can occur either intraoperatively or postoperatively, and can involve the femur, tibia, or patella.
Intraoperative fractures can potentially occur at any stage of the TKA procedure, including exposure, bone preparation, placement of trial components, insertion of the final components, and insertion of the polyethylene insert.
Intraoperative femur fractures can occur in either the diaphyseal or metaphyseal region. Metaphyseal fractures are more common and can include the medial or lateral condyle, medial or lateral epicondyle, and the supracondylar metadiaphyseal region. These fractures occur most commonly during bone preparation, trialing, and implantation of final components. Technical errors that can lead to these fractures include inadequate preparation of the femoral box in posterior-stabilized (PS) designs and aggressive impaction of the femoral trial or final component in osteopenic patients. In one series, all intraoperative femoral fractures occurred in females with narrow femurs in which a PS implant was used. Another series reported 40 distal femur intercondylar fractures in 898 primary PS TKAs. The authors attributed this to a wide box cut in their implant. Their fracture incidence decreased once they changed their implant to one with a smaller box cut. Alden et al. reported a relative risk of 4.74 when using a PS design as opposed to a cruciate-retaining (CR) design. It is important that surgeons understand the subtleties of the implant that they are using and that meticulous attention be paid to deciding the medial-lateral placement of the box cut to avoid this complication.
Diaphyseal fractures are less common and typically involve penetration of the anterior or posterior cortex from a malpositioned intramedullary drill or guide during distal femoral preparation ( Fig. 11.1 ). These fractures are easily missed intraoperatively because they typically will not cause instability or deformity and the femoral diaphysis is not exposed during the procedure. ,
Intraoperative tibia fractures are less common than femur fractures. These fractures can occur during any stage of the procedure and can include the medial or lateral plateau; the tibial tubercle; and the medial, lateral, anterior, or posterior cortex. , These fractures are more commonly encountered during revision surgery with forceful extraction of a well-fixed component, during cement removal, aggressive impaction of the tibial component, eccentric preparation or placement of a tibial stem, and cone or sleeve preparation and impaction. ,
Supracondylar femoral fractures are the most common postoperative fractures around total knee implants. , These fractures are often the result of low-energy trauma such as ground-level falls, although they also occur after higher-energy trauma, seizures, and manipulation of a stiff knee.
Patella fractures are the second most common periprosthetic fracture after primary TKA. Their etiology can be either traumatic or atraumatic. They can occur after severe thinning of the native patella due to the disease process, and can be associated with avascular necrosis of the remaining bone. A systematic review reported that only 11.6% of fractures had a traumatic incident, with the remainder occurring without any significant trauma. Most of these fractures occur within 2 years of the index procedure. These can be very difficult fractures to treat, as the index procedure can decrease patellar blood supply and bone stock. It can also be difficult for patients to return to satisfactory function with treatment, especially if the extensor mechanism is disrupted or does not return to baseline.
Periprosthetic tibia fractures are the least frequent postoperative periprosthetic knee fracture. These fractures previously occurred more frequently with earlier total knee designs. However, these fractures occur more rarely now with the use of short-stemmed or keeled tibial baseplate designs. , Interestingly, a systematic review showed that nearly 80% of fractures of the tibial plateau did not have a traumatic etiology.
Various classification systems have been proposed for distal femur fractures. Neer and associates reported the first classification system in 1967, which is based on the amount and direction of displacement of the fracture. Rorabeck and Taylor proposed a classification system 30 years later that added the fixation of the prosthesis, with Types I and II being nondisplaced and displaced fractures, respectively, with a well-fixed component, while Type III fractures involve a loose component and are either displaced or nondisplaced ( Table 11.1 ). In 2006, Su and associates proposed a classification system based on the location of the fracture, with Type I proximal to the prosthesis, Type II starting at the anterior flange of the femoral component and extending proximally, and Type III fractures involving any other aspect of the femoral component distal to the most proximal aspect of the anterior flange ( Fig. 11.2 ). Kim et al. proposed a classification system that takes into account the stability of the implant, the volume and quality of bone stock available for fixation, and the reducibility of the fracture.
|Displaced or nondisplaced
The most commonly used classification for patella fractures was described by Ortiguera and Berry. This classification system considers both the stability of the patellar implant and the integrity of the extensor mechanism. Types I and III fractures have an intact extensor mechanism with stable and loose patellar components, respectively. Type II fractures have a disrupted extensor mechanism ( Table 11.2 ).
|Stable or Loose
With respect to periprosthetic tibia fractures, Felix et al. proposed a classification system for periprosthetic tibial fractures. This system divided fractures into Types I to IV based on anatomic location (I: tibial plateau; II: adjacent to stem; III: distal to prosthesis; IV: tibial tubercle) and into subcategories based on prosthesis fixation (A: well-fixed; B: loose; C: intraoperative; Fig. 11.3 ).
The incidence of all periprosthetic knee fractures is likely underreported due to clinically insignificant fractures going unnoticed. The number of these fractures has been rising and is expected to continue to increase. This is due to both the rising numbers of primary TKA and the increasing age of the population with a higher rate of osteoporosis and comorbidities contributing to a higher predilection for suffering falls.
The true incidence of intraoperative fractures is likely underreported, as a clinically insignificant fracture may be missed. The incidence of intraoperative fractures varies in the literature. As stated earlier, Lomardi et al. reported an incidence of 4.4% using an implant with a wider box, and this rate was reduced to 0.2% when a different implant was chosen. Alden et al. reported an incidence of any intraoperative fracture at 0.39%, with 73.1% of those being femur fractures.
The incidence of postoperative distal femoral periprosthetic fractures has been reported to range between 0.3% and 2.5% of all knee implants and has been reported as high as 38% in revision implants. , Ebraheim et al. reported in a systematic review that Rorabeck Type II fractures were the most common type documented.
The tibia is less commonly involved in periprosthetic knee fractures, with a reported incidence of 0.07% to 1.7% after primary arthroplasty and 0.36% after revision arthroplasty. , , Most fractures that occur are Type I, followed by Types II and III. Type IV fractures are quite rare, with a systematic review reporting only 2 in 112 periprosthetic fractures being Type IV. Intraoperative fractures are more common in revision surgery, with one series reporting a rate of 4.9% of revision knee arthroplasty cases when using press-fit stems.
The incidence of patella fracture after TKA varies widely and can depend on whether the patella was resurfaced or not. The incidence of fracture in the unresurfaced patella is reported to be 0.05% of cases, while the rate in resurfaced patellas is reported between 0.2% and 21% of cases. ,
Risk factors reported for distal femur periprosthetic fractures include female sex, osteoporosis, neuromuscular disease, cognitive disorders, chronic steroid use, inflammatory arthritis, infection, and polyethylene wear resulting in osteolysis. Increased constraint can predispose to distal femur fractures, as this can cause an increased stress on the bone-implant interface with the implant having a much higher modulus of elasticity. , , Additionally, constrained components require a larger box cut, which results in a decreased segment of bone to bridge the medial or lateral condyle to the femoral diaphysis and to each other. Anterior femoral notching, which occurs when the anterior flange of the femoral component violates all or part of the anterior cortex of the femur, has an associated theoretical increased risk for distal femur periprosthetic fracture. Biomechanical studies have shown decreased bending and torsional loads to failure when notching is present. , However, clinical studies have not shown an increased rate of fracture with notching.
The risk of postoperative tibia fracture can be increased when the baseplate is placed in varus or with a loose component. Intraoperative fractures are typically due to technical error, most commonly when the baseplate is impacted too aggressively. Intraoperative fracture risk is also increased in revision surgery and when a tibial tubercle osteotomy is used to improve exposure. ,
Periprosthetic patella fractures can occur in both resurfaced and nonresurfaced patella, and literature has shown a higher rate of fracture in resurfaced patellas. , Bourne and Burnett classified the risk factors for patella fractures into three groups. The first group relates to patient factors, the second relates to implant design issues, and the third includes intraoperative factors related to technical error. Patient factors include osteoporosis, male sex, inflammatory arthropathy, and excessive postoperative mobilization. Factors related to implant design include cementless implants and implants with a single large central peg. Technical errors are the most preventable and include aggressive use of patellar clamps, asymmetric or over-resection of the patella, and femoral implant malrotation. Femoral malrotation leads to patellar instability and insult, which can increase the risk for fracture. The surgeon should also take care to preserve the blood supply to the patella, as ischemia increases risk for fracture. It has been shown that routine exposure during TKA causes insult to the patellar blood supply. Some authors suggest that the surgeon should retract the patella laterally rather than everting. One study used Doppler flowmetry to quantify patellar ischemia and showed that eversion resulted in a much higher degree of ischemia than lateral retraction alone. Additional factors that can reduce patellar blood supply include aggressive resection of the fat pad and lateral retinacular release. Residual bone thickness is also a significant factor that can lead to fracture. The precise appropriate thickness is being debated in the literature. However, residual bone thickness of less than 15 mm may predispose toward a patella fracture.
Intraoperative Periprosthetic Femur Fractures
As indicated earlier, intraoperative femur fractures can occur at any time during the procedure and can be metaphyseal or diaphyseal. The variety of fracture locations and patient factors makes it difficult to create a specific treatment algorithm. Treatment is typically individualized to the patient based on surgeon experience, fracture pattern, and bone quality. Treatment options include nonoperative observation, open reduction and internal fixation, and revision. Fixation strategies that have been described include plate-and-screw fixations ( Fig. 11.4 ), stemmed prosthesis, screw fixation, suturing, bone grafting, and cementing. , Principles of management are based on obtaining a construct that can withstand shear, axial, and rotational forces once a final component is implanted. The most common fractures are single condylar fractures, as discussed earlier. If nondisplaced, these fractures often have an intact periosteum and can be treated without additional fixation and protected weight bearing. Displaced condylar fractures can be treated with screws with or without washers in most circumstances or a plate-and-screw construct. The use of stout, nonabsorbable suture has also been described. The surgeon can also choose to bypass the fracture with the addition of a stem to the femoral component. , Intraoperative diaphyseal fractures should be treated with a long-stemmed prosthesis. The stem should bypass the fracture by at least two cortical diameters. All patients with intraoperative femoral fractures should be put on protected weight-bearing precautions until the fracture is healed.
Postoperative Periprosthetic Femur Fractures
Nonoperative treatment is indicated in cases of stable fractures with minimal displacement and a well-fixed and well-aligned component. This type of fracture is somewhat rare, as most fractures are either displaced or unstable. Stable fractures can be treated with cast or brace immobilization, followed by progressive range of motion exercises. There is no agreed upon duration of immobilization or weight-bearing restrictions in the literature. Moran et al. treated Type I fractures with either cast immobilization or a hinged brace for 4 to 6 weeks with partial weight bearing, and treated Type II fractures with cast immobilization for 6 weeks with non–weight bearing. All Type I fractures went onto union; however, none of the Type II fractures healed with a satisfactory result. Patients treated nonoperatively must be monitored closely with frequent radiographs to ensure that their fracture does not displace and go on to malunion or nonunion ( Fig. 11.5 ).