Introduction

Forty years ago, knee arthroplasties were introduced with one or more femoral and tibial sizes. Gradually, left and right components as well as more sizes were added to better match smaller and larger knees in different patients. The components’ shape and dimensions were optimized using anthropometric data, and asymmetric femoral and tibial components were introduced. However, to this day there is no specific implant for the lateral tibiofemoral joint, on the femoral and on the tibial side. Common practice is to use a left medial component for the right lateral and a right medial for the left lateral tibial plateau. One would imagine that, besides anteroposterior (AP) and mediolateral (ML) dimensions and their ratio, different shapes or different curvatures would have been introduced to optimize each patient’s anatomy and to restore closer to normal knee kinematics. However, this is not the case. Shapes are determined by manufacturers and come in different magnifications of their geometries. This is done for both femoral and tibial components. On the femoral side, each manufacturer offers one specific J-curve with different magnifications, matching the condylar geometry to some degree. One manufacturer even offers a single-radius femoral component. The question remains whether a more anatomic reconstruction of the joint surfaces would further enhance clinical outcome. More sophisticated measures are needed, such as measurement of proprioception, muscle function, and gait analysis to reflect the changing demographics of younger, more active patients considering knee replacement surgery.

The selection of a specific off-the-shelf implant is driven by a surgeon’s familiarity with surgical technique and implants, quality of sales support, and ease of a surgical technique. Off-the-shelf implants lack the ability to best fit for an individual surface geometry. Pricing has and will play a more important role in the future. An implant should not be chosen for any other reason besides it being the best fit for a patient.

Femoral Components

The overall goal of selecting a femoral component is optimal fit. The femoral component should not overhang, but it should be wide enough and match the geometry without impingement of the medial spine or the posterior cruciate ligament. The transition to the femoropatellar joint should be smooth, not prominent, and the femoral component should not be too long, specifically on the lateral side to avoid patellofemoral (PF) impingement. Hernigou and Deschamps¹ reported on 99 unicompartmental arthroplasties (UKAs) with a mean follow-up of 14 years. Of these, 29% were symptomatic due to patellar impingement through the femoral component. In this group, lower Knee Society scores were related to problems climbing stairs. They proposed two failure mechanisms; one may be related to an over-resection off the posterior condyle, which consequently moves the femoral component more anteriorly. They also hypothesized that poor intraoperative judgment of the linea terminalis, which describes the trochleocondylar junction, could lead to oversizing the femoral component. Figure 15–1 shows the transition between the medial and lateral condyles to the trochlea. The linea terminalis is the result of the medial and lateral meniscal impression on the femur and can be seen on lateral radiographs. In Herningou and Deschamps’s series, the trochleocondylar junction was an average of 3 mm anterior on the medial and 4 mm posterior on the lateral side relative to a line representing the slope of the intercondylar roof.¹ Berger et al.² reported the results of 59 UKAs with a follow-up of 11–15 years. The main failure mode was symptomatic patellofemoral osteoarthritis in 10%; this was observed despite a pristine PF joint preoperatively. In a group of 48 living patients who had an average radiographic follow-up of 17 years, Squire et al. reported evidence of PF osteoarthritis in 35 patients (87%).³ A total of 14 knees were revised. Six patients were revised for disease progression, but it was not defined whether this was related to PF osteoarthritis.

Figure 15–1 The linea terminalis is more anterior medially as compared to the lateral side.

The linea terminalis has to be respected to avoid PF impingement. Both surgical technique and femoral implant design play a role. The distal femoral cut in most systems is not based on an intramedullary rod, but solely on the tibial cut. It is difficult for the surgeon to judge the correct amount of flexion of the distal femoral cut. This step determines the amount of flexion in which femoral component is implanted. If the knee is hyperextended, the distal femoral cutting block cuts in more extension. If the knee is held in more flexion, the component results in a more flexed position. However, the tibial cut can vary depending on whether there is more or less slope, which additionally contributes to more or less flexion or extension. Individualized femoral jigs may help to resolve this source of surgical error. Figure 15–2 shows a femoral component in too much flexion, resulting in PF impingement. All three studies cited above reported increasing PF osteoarthritis in the second decade. All studies used an only-femoral component (Fig. 15–3), which means that the anterior edge is not curved below the subchondral bone. This is in contrast to a recessed femoral component where the edge of the implant extends below the subchondral bone as shown in Figure 15–4. Using a recessed single-radius femoral component, progression of PF osteoarthritis was not seen in a series with a follow-up between 10 and 20 years.⁴ It remains unclear, since there is no reported 10- to 20-year follow-up series of a tapered J-curve design, whether this is related to the recesses design itself or to the single-radius design. The single-radius femoral implant restores the posterior condylar offset by balancing the flexion gap prior to the extension gap, so the surgical technique is very focused on restoring the posterior condyle. Additionally, its shorter design, replacing only the posterior condyle, may increase femoropatellar offset, which may result in decreasing PF stresses and thereby reducing the likelihood of PF osteoarthritis. However, since PF peak stresses are lowest in extension and increase during flexion,^5,6 J-curve femoral components reconstructing the posterior condyle should achieve the same offset compared to a single-radius design (see Fig. 15–4). Restoring the cylindrical posterior condyle,⁷ therefore, should not result in significantly different PF stresses.

Figure 15–2 Only-femoral designs are prone to surgical errors. Too much flexion of the femoral component can lead to patellofemoral impingement.

Figure 15–3 Only-femoral designed components rely on a distal femoral bone cut. The edge of the femoral component thins, but does not go below the subchondral bone.

Figure 15–4 J-curve and single-radius designed femoral components are available where the edge of the component extends below the subchondral bone.

Only-femoral designs may be more susceptible to PF osteoarthritis as a failure mode in the second decade due to impingement between the patella and the edge of the femoral implant. This may be additionally enhanced with the decreasing of cartilage thicknesses as we get older. Individualized UKAs have several advantages compared to off-the-shelf implants: computed tomography (CT) data are utilized to identify the linea terminalis and the implant is designed with respect to the transition to the PF joint with short lateral and longer medial components. The length of the components is related to the restoration of the posterior condyle. A surgical plan provides detailed information on how much bone needs to be resected off the posterior condyle. An example of 6 mm of posterior condylar resection is shown in Figure 15–5. Therefore, an incidental too-anterior placement of the femoral component is addressed by preoperative planning, individualized femoral cutting blocks, and cross-checking through the correct amount of resected bone. In order to address the problem of PF impingement, the anterior edge is recessed below the subchondral bone. Figure 15–6 shows a three-dimensional model demonstrating this design feature.

Figure 15–5 A preoperative surgical plan provides detailed information on how much bone needs to be resected off the posterior condyle. A specific instrument (L-guide, in blue) facilitates appropriate posterior resection. The instrumentation is individualized and single use.

Figure 15–6 The anterior tip of the femoral component is designed to recess underneath the subchondral bone to address femoropatellar impingement.