Numerous devices have been described to assess soft-tissue tension during TKA. We use a tensioner made up of a single fixed tibial plate which sits on the tibial cut surface and two separate moveable plates for the medial and the lateral femoral condyles (Fig. 2.1a). These two femoral plates can be moved independently using a screw mechanism to increase or decrease the tension medially and laterally. In full extension, the aim while assessing ligament tension is to achieve a limb alignment as close to neutral as possible and equal tension both medially and laterally as assessed by the surgeon (Fig. 2.1b). The final stability is checked in full extension using spacer blocks where a valgus and a varus stress are applied to determine the degree of laxity (Fig. 2.1c). This step is again repeated in 90° flexion using the tensioning device and also using a spacer block with the AP cutting block in place (Fig. 2.1d, e). The aim is not only to ensure that there is optimum stability in full extension and 90° flexion but also to confirm that any opening on stressing is equal on both sides and which also springs back equally.

Based on the type of knee deformity and the degree of initial soft-tissue imbalance, the site and the structures to be released vary from knee to knee. A thorough knowledge of what structures can be released during TKA for any given deformity is essential. For a varus knee, no release on the lateral side should be performed, and for a valgus knee the opposite is true. Similarly, for a knee with flexion deformity, posterior release is necessary, whereas in a hyperextending knee it should be strictly avoided. We completely avoid performing any sort of release of the collaterals and avoid completely releasing the pes anserinus or dividing the popliteus tendon. Most deformities and imbalance can be corrected releasing capsular structures and adhesions rather than the collaterals and supplementing these releases with osteotomies. It must be pointed out that the collateral ligaments themselves do not undergo contracture/shortening.

We follow a graduated, controlled and stepwise approach to soft-tissue release during TKA. Removal of osteophytes as the initial step considerably improves deformity and decreases the quantum of soft-tissue release required. A preliminary soft-tissue release is performed as part of exposure and following osteophyte excision. Rigid deformities may not be fully corrected with a preliminary release. Similarly, although soft-tissue release may achieve full correction of deformity, it may fail to achieve medio-lateral balance. Such cases with residual deformity or medio-lateral imbalance will require further, extended releases supplemented by a reduction osteotomy [14].

With the knee in full extension, the tibial and the distal femoral resection are made perpendicular to their respective mechanical axes (Fig. 2.2a). The aim is to create a rectangular extension gap equal on both medial and lateral side. However, despite accurate bone cuts, the extension gap may be unequal or quadrilateral owing to medio-lateral soft-tissue imbalance as determined by spacer blocks (Fig. 2.2b). This disparity needs to be rectified using appropriate soft-tissue releases to restore limb alignment and medio-lateral balance (Fig. 2.2c). The extent of medio-lateral soft-tissue disparity in extension after the bone cuts can be predicted using the joint divergence angle on preoperative full-length radiographs (see Chap.​ 1).

With the knee in 90° flexion, the tibial and posterior femoral cuts should generally be parallel to the epicondylar axis and perpendicular to the anteroposterior axis of the femur (Fig. 2.3a). We mark the level of the tentative posterior femoral cut after assessing flexion gap balance using the tensioner (Fig. 2.3b, c). However, it is very important to achieve good medio-lateral balance in flexion before determining the level and performing the posterior femoral cut. In varus knees, if the posterior femur is cut in the presence of excessive medial tightness, then the femoral component may get placed in excessive external rotation and vice versa (Fig. 2.3d–e). Similar to when the knee is in full extension, the aim at 90° knee flexion is to create a rectangular flexion gap equal on both medial and lateral side. Hence, the determinants of flexion gap include distal femoral and tibial cuts, medio-lateral soft-tissue tension in 90° flexion, femoral component size and position and the tibial insert thickness.

Based on our technique of creating the extension gap first, the way to flexion-extension gap balancing involves creating a flexion gap matching in dimension to the previously created extension gap. The key to achieving this is the size and position of the femoral component. Using the method previously described, the flexion gap is assessed using a spacer block with the AP cutting block of the estimated femoral component size pinned in place. This must have no slots posteriorly. Using a stylus or “angel wing”, we ensure that there is no notching. If the flexion gap is tighter for the corresponding spacer block used previously in the assessment of extension gap, the femoral component needs to be undersized or moved anteriorly in order to equalise the flexion-extension gaps. Slight flexion or posterior shift of the femoral component may also be helpful to tackle a flexion gap that exceeds the extension gap. However, the surgeon needs to be cautious in not overdoing this and causing either anterior notching or post impingement against the box in extension. The AP block is placed accurately before performing the anterior and posterior femoral cuts by using “combined referencing” where multiple references are used to ascertain the position of the AP cutting block and subsequently the femoral component (Fig. 2.4).