Infection must be considered in the evaluation of every patient with a painful total knee replacement. It is a most devastating and feared complication that often threatens the function of the joint, the preservation of the limb, and the health of the patient. Deep infections occur in 0.39–3.9% of primary total knee replacements and, on average, three times higher in revision cases [6, 7]. Body mass index ≥35, diabetes mellitus, male sex, American Society of Anesthesiologists (ASA) score ≥ 3, diagnosis of osteonecrosis, and a diagnosis of posttraumatic arthritis have been shown to increase the relative risk of deep infection [8]. The most common organisms are Staphylococcus aureus and Staphylococcus epidermidis. Methicillin- and vancomycin-resistant organisms have become increasingly prevalent and difficult to treat. The diagnosis of infection should start with a thorough history and physical examination. Persistent pain is the only consistent finding with infection, although a draining wound or history of wound problems or any erythema must also raise the suspicion for infection (Fig. 3.1) [9]. Serum studies including white blood cell count, erythrocyte sedimentation rate, and C-reactive protein are useful, particularly in following the course of treatment. In patients undergoing revision knee arthroplasty, erythrocyte sedimentation rate >30 mm/h has a 63% sensitivity and 55% specificity for infection, whereas C-reactive protein >10 mg/L had a 60% sensitivity and 63% specificity [10]. Bone scans are also helpful, with sensitivities and specificities of approximately 84% [11]. Aspiration of the knee should be performed, and the fluid should be analyzed for cell count with differential and culture. Cell count and neutrophil differential both below a cutoff value of >1100 cells and >64%, respectively, yield a negative predictive value of 98.2% [12]. However, the existing diagnostic criteria for periprosthetic joint infection in the literature vary widely, and even when the results of the aspirate are combined with serum inflammatory markers, there remains a large variance in sensitivity (54%–100%) and specificity (39%–100%) [13]. Finally, tissue taken intraoperatively may be sent for frozen section pathological examination. Greater than ten polymorphonuclear leukocytes per high-power field is implicated in infection with a sensitivity of 84% and a specificity of 99% [14]. Hence, the diagnosis of infection must be made based on careful history and physical examination using all available data, rather than basing the diagnosis on one particular test.

Treatment of a total knee infection is often based on the timing and duration of the infection as well as the implicated organism and the status of patient’s overall health. Decisions must then be made whether to attempt prosthesis retention, one-stage exchange, or two-stage exchange. A glycocalyx layer formed around the prosthesis may prevent antibiotic penetration to the prosthesis, rendering antibiotic treatment alone ineffective. Surgical treatment remains the mainstay. Aggressive treatment for superficial wound infections is recommended, as many of these infections actually involve deeper tissues. Primary debridement within 10 days of symptom onset has a reported success rate of 56% in patients with low-grade organisms, but the success rate is diminished to 8% in the presence of Staphylococcus aureus infections [15]. Even lower rates of success are reported for using this approach for chronic infections. Arthroscopic debridement has only seen moderate success in the eradication of acute (within 4 weeks of surgery) infections, providing eradication in 52% of patients [16].

Prosthetic exchange is the primary mode of treatment when eradication of the infection is the goal. Single-stage exchange may be considered when an acute infection with a relatively low-virulence gram-positive infection is encountered in a competent host. One study showed 89.2% success with single-stage exchange in which there were gram-positive infection, absence of sinus tract, antibiotic-impregnated cement in the new prosthesis, and 12 weeks of adjuvant antibiotic treatment [16]. The most widely accepted approach, however, is the two-stage exchange in which aggressive irrigation, debridement, synovectomy, and prosthesis removal are performed, followed by reimplantation after a period of intravenous antibiotics. During the interim, a spacer of antibiotic-impregnated methyl methacrylate is often used. With this technique, overall infection-free survivorship was shown to be 85% at 5 years and 78% at 10 years [ 17]. Up to 97% eradication rates are reported with this technique [12]. The use of a PROSTALAC functional spacer made of antibiotic-laden cement with a small metal-on-polyethylene articulation is of interest because of its potential for enhanced function and maintenance of good alignment and stability of the knee. This facilitates second-stage procedures. Using this technique in a two-stage exchange with a mean 4-year follow-up, cure rates of 91% have been demonstrated [18]. Although this is promising, further outcome-based studies are necessary.

It is critical to always maintain a high index of suspicion for infection and to treat infections aggressively. All painful total knee replacements must be evaluated for the possibility of an indolent infection.

Polyethylene wear in total knee arthroplasty continues to affect the longevity of modern total knee replacements. Wear and aseptic loosening have been shown to be the second most common modes of failure requiring revision surgery in the United States, accounting for up to 16.1% of revision operations [31]. From a basic science standpoint, osteolysis is the granulomatous response to polyethylene, polymethyl methacrylate, and metal debris, which are formed by both the articulating and nonarticulating (undersurface) surfaces of the prosthetic knee. Delamination, adhesion, and abrasion cause the liberation of loose particles that contribute to osteolysis. Osteolysis was implicated in 2% of early and 9% of late failures of total knee arthroplasties requiring revision surgery [32]. Risk factors include incongruent articulations, poor tibial locking mechanisms, thin polyethylene, sterilization of polyethylene with gamma irradiation in air, fixation screws in the tibial base plate, and an extended shelf life of the polyethylene implants. Most patients remain asymptomatic. However, some patients have a boggy synovitis and mild to moderate pain with activity. A triad of effusion, pain, and change in coronal alignment, usually into varus, is strongly suggestive of accelerated polyethylene wear. Identification of a lytic osseous defect, absence of bone trabeculae, and geographic demarcation makes the diagnosis radiographically (Fig. 3.5). The presence of the components may obscure the lesions on radiography, particularly as they are most commonly found within 2 mm of the tibial component and in the posterior femoral condyles. If osteolysis is suspected, computed tomography is a useful tool to evaluate the size of the osteolytic lesion [33]. Nuclear medicine studies may also demonstrate increased uptake around loose components. Osteolysis must be distinguished from radiolucent lines that are a common finding in radiographic surveillance of total knees. Lysis requires a complete radiolucent line of greater than 2 mm in length. Smaller lines are of unknown significance and may be followed clinically. Ranawat et al. noted radiolucent lines in 72% of the tibiae, 54% of the femurs, and 33% of patellae [3]. Not all of these represented osteolysis. Treatment of these lesions primarily depends on whether the osteolysis is associated with loose prosthetic components. It is essential to review serial radiographs to determine if radiolucent lines are progressive. Well-fixed components with lytic lesions may be treated with exchange of the polyethylene insert and bone grafting of the lesions. However, isolated tibial insert exchange resulted in a 63.5% cumulative survival rate at 5.5 years [34]. They recommended that limited revision of the polyethylene should be avoided if severe delamination is present, if there is significant undersurface wear of the polyethylene suggesting an inadequate locking mechanism, and if there is early failure within 10 years of the index operation. Revision of loose components with bone graft is indicated for lysis associated with loose components. It is important to have a full complement of revision instruments available with stems, wedges, and allograft when performing these revisions, as radiographs not only underestimate lesion size but do not take into account bone loss with explanation of the loose components (Fig. 3.6).

Symptomatic axial instability of a total knee arthroplasty, including valgus-varus and flexion-extension instability, is a potential cause for pain and disability following total knee replacement. It occurs in 1–2% of patients and may be present in either posterior stabilized or cruciate-retaining knees. Overall, instability accounts for 10–20% of all total knee revisions, following only infection and aseptic loosening in prevalence [35]. Instability may be caused by trauma, ligamentous stretch, inadequate balance at the time of surgery, or a systemic disorder such as Ehlers-Danlos disease.