Anteroposterior radiographs should be scrutinized for gross limb alignment, component sizing, and positioning. Implant loosening in a later stage as well as failure of the TKR might also be identified (Fig. 28.1). In addition, radiographs give information on periprosthetic radiolucencies representing osteolyses, fractures, excessive polyethylene wear, or heterotopic ossifications (Fig. 28.2).

In early failure of TKR, unambiguous component malpositioning may be one of the most important pathologies to be detected on conventional radiographs. Component alignment in the coronal plane has been considered to be the most important factor in determining long-term survival of TKR [1]. Although we should be aware that the optimal method of TKR component position is by 3D CT, component position analysis on radiographs should receive special attention, as it is more commonly available [2]. In the coronal plane, the position of the tibial and femoral TKR components is evaluated in relation to the anatomical shaft axes of femur and tibia, respectively. It is usually depicted as lateral distal femoral angle (LDFA) and medial proximal tibial angle (MPTA) (Fig. 28.3a, b). In general, the femoral component is placed in 6° of valgus, whereas on the tibial side a perpendicular position is pursued. However, standard reference values and variability of these angles are under ongoing discussions. The position is dependent on the specific implant and preexisting individual patient’s anatomy [3–6]. For a proper analysis of a painful TKR, meticulous evaluation of the preoperative radiographs is also essential (Fig. 28.4).

The rotational position of the femoral and tibial TKR components can be best assessed on 3D reconstructed CT images. Kanekasu et al. [7] proposed a method to determine femoral component rotation on an axial view of the distal femur. This radiograph is taken in 90° of flexion. The angle measured between the transepicondylar axis and the posterior condylar axis showed an acceptable correlation with the rotational position analyzed on CT scans. However, tibial rotation cannot be assessed on these radiographs.

The assessment of the joint line height helps to understand the kinematics of the knee. The joint line can be approximated on the basis of the fibular head, the femoral epicondyles, or the superior pole of the patella in relation to the superior aspect of the trochlear groove. Again, available preoperative radiographs or radiographs of the native contralateral knee help to determine the anatomical location of the joint line [8] (Fig. 28.5).

Sizing and positioning of the TKR components in the mediolateral dimension can be assessed more easily for the tibial component than for the femoral one. A medial overhang of the tibial component is usually more acceptable than on the lateral side. On the other hand, if the implant size is too small or the TKR component is not supported by cortical bone, it might be prone to subsidence (Fig. 28.6). This risk is multiplied if a medially undersized tibial TKR component is accompanied by varus malalignment.

Focal periprosthetic osteolysis is usually attributed to polyethylene wear particles but can also be found in cases of infection. The most prominent signs are radiolucent lines. Osteolytic lesions usually develop adjacent to the cement-bone interface in cemented components or implant-bone interface in cementless fixation. Depending on their size and progression, they destabilize the implant and may result in considerable loss of periprosthetic bone.

The presence, localization, and extent of periprosthetic bone loss guide the strategy for revision surgery and therefore have to be carefully evaluated. The most widely accepted classification of periprosthetic bone loss is that of the Anderson Orthopaedic Research Institute (AORI), where bone loss is categorized in three groups for femur and tibia, respectively [9] (Fig. 28.7):

Gross instability can also be seen on weight-bearing radiographs. However, for more subtle cases, please see Chap.​ 32.

For adequate evaluation of the mechanical alignment, a long-leg view is recommended. It has been shown that femoral and tibial TKR component positioning in the coronal plane can be analyzed on a knee-only radiograph with comparable accuracy as on long-leg views, but evaluation of the mechanical axis is limited to a long-leg radiograph [10] (Fig. 28.5).