The lower extremity from an in utero position of internal rotation slowly undergoes lateral or external rotation throughout the growth phase of the child till skeletal maturity [9]. Staheli and Engel [10] in their study of tibial torsion in children estimated tibial torsion at 15° of external rotation at skeletal maturity. Based on the orientation of the bimalleolar axis with respect to the flexion-extension axis of the knee joint, the tibia shows dynamic internal rotation when the knee is flexed [11, 12]. During gait, the tibia shows dynamic internal rotation with respect to the femur at mid and terminal stance and dynamic external rotation from toe off until slightly before heel strike [12]. In normal subjects, the tibia is typically in external rotation with Caucasian limbs showing greater external rotation compared to Asian limbs [13].

Whether abnormal torsion is a cause or a consequence of knee osteoarthritis is still unclear. Nagamine et al. [14] in their CT scan analysis of normal and osteoarthritic tibiae in Japanese subjects had suggested that the traditional way of sitting on the floor with the foot internally rotated (tatami position) since childhood may cause growth disturbance at the proximal tibial meta physis, causing tibia vara and medial torsion of the tibia. Krackow et al. [15] in a recent study to quantify the association between medial knee loading and tibial torsion reported that subjects with medial OA and associated tibial intorsion walked with significantly greater knee loading compared to controls. Hence, the incidence and pattern of torsional deformity in patients with knee OA shows wide variation depending on the ethnicity of the population and severity of deformity.

Several techniques are used to determine the rotational position of the femoral component during TKA. These include using bony landmarks on the distal femur, soft-tissue tension and flexion gap (gap balancing) and computer navigation [16]. However, no single technique has emerged as a clear winner [17], and most surgeons use a combination of two or more techniques. Distal femoral bony landmarks are most frequently and principally used to determine rotation of the femoral component (Fig. 7.1). However, several reports have underlined the fact that such bony landmarks show wide variation not only among arthritic knees but even among normal subjects [2, 18–22]. Furthermore, use of one specific bony landmark may be unreliable to determine femoral component rotation and may be affected by severity and type of underlying arthritic knee deformity. For example, the anteroposterior axis (Whiteside’s line) may not be reliable in cases with severe patellofemoral arthritis which may cause significant trochlear wear and distortion [23, 24], and the posterior condylar angle (PCA) may not be reliable in knees with severe wear and bone loss of the posterior condyles [25]. Hence, in such cases some other technique must be used to determine femoral rotation. The gap-balancing technique involves rotating the femoral component so as to achieve medio-lateral flexion gap symmetry (with the help of a tensioning device such as laminar spreaders) after having achieved a balanced and symmetric extension gap. However, the femoral rotation required may vary depending on the amount of distal femoral bony loss or distortion and the extent of medio-lateral flexion gap asymmetry. Although computer navigation has proven to improve limb and component alignment during TKA, studies have reported malrotation of the femoral component during CAS TKA when a single anatomic landmark is used as reference [26, 27]. For example, using only the epicondylar axis as reference may lead to excessive external rotation of the femoral component. However, using a combination of anatomic and kinematic data has been reported to improve accuracy of rotational alignment of the femoral component [27].

On the tibial side, options to determine tibial component rotation include using bony landmarks and putting the knee through an arc of flexion and extension and positioning the tibial tray vis-à-vis the femoral component (the “self-positioning” method) [16]. Traditionally, the tibial tubercle (junction of the middle and medial one-third) has been used as a fixed bony landmark to determine the rotational alignment of the tibial tray. However, similar to distal femoral bony landmarks, the tibial tubercle may show wide variations among patients depending on the severity of knee deformity and ethnicity [2, 28–30]. Sun et al. [29] in a CT scan-based study of Chinese osteoarthritic knees reported that the tibial component has a tendency to be placed in greater external rotation when the medial one-third of the tibial tuberosity is used as a landmark in arthritic knees with varus or valgus deformities. Hence, they concluded that the tibial tuberosity is an unreliable rotational landmark for tibial tray in Asian patients with deformed, arthritic knees [29]. In a similar study on European subjects, Bonnin et al. [31] reported significant variation in the position of the tibial tuberosity which caused not only excessive external rotation of the tibial component but also deficient coverage of the tibial cut surface by the tibial tray.