Planning for surgery and avoiding potentially catastrophic skin necrosis require an awareness of the previous incisions used at previous knee surgery. A laterally based incision from the previous HTO should not provide significant difficulties as long as a skin bridge of at least 8 cm can be achieved. This may require a slightly medially based skin incision rather than a true midline incision. A previous transverse incision should pose no problem as long as the new incision is perpendicular to it. Where parallel incisions are present, the more lateral incision is recommended, as the blood supply to the extensor surface is medially dominant. Very rarely, a sham incision can be used before the definitive surgery, to more safely assess the potential wound healing. Jackson et al. [25] noted a 30% rate of primary wound healing in TKA after failed HTO, with a 20% incidence of deep infection (Fig. 20.2).

Scar tissue between the patellar tendon and the proximal anterior tibia often makes eversion of the patella after a previous HTO more difficult [22]. Release of this scar tissue and excision of a thickened fat pad can improve exposure. The patellofemoral ligament should be routinely released to improve lateral exposure. Meding et al. [2] reported that this was adequate to complete the tibial exposure in each case.

If difficulty with exposure is still encountered, then an early lateral release can be performed [8, 21]. Personal experience of the senior author (RDS) in 74 consecutive conversions of failed HTO to TKA is of a lateral release rate of 38% compared with a 30% lateral release rate in 1000 consecutive arthroplasties from the same era. Nizard et al. [22] reported a lateral release rate of 24% in their post-osteotomy group compared with just 2% in their control group. If exposure is still compromised, then a quadriceps snip is recommended. A tibial tubercle osteotomy should rarely be required for exposure, although Nizard et al. [22] used a tibial tubercle osteotomy in 7 of 63 post-osteotomy cases. Finally, a pin through the patella tendon insertion intraoperatively is strongly recommended, as a prophylactic measure to protect it from avulsing (Fig. 20.3).

Various fixation devices are usually used in HTO. Options include staples, a compression plate and screws, a blade plate, and other similar hardware. Preoperative planning is required to assess whether the hardware will interfere with the TKA (Figs. 20.4 and 20.5). If not, then the HTO fixation device does not require removal unless its presence is symptomatic to the patient.

If the hardware will interfere with the tibial jigs or implant, then the decision as to whether to perform the TKA in one stage or two would depend on whether a separate incision is required for hardware removal, the size and placement of the hardware, and the site of previous incisions.

For 2-stage arthroplasty, an interval of 6–12 weeks after hardware removal should be used to enable good wound healing before the TKA. Also, cultures of the osteotomy site should always be obtained at the first-stage procedure.

Patella infera is often seen after a closing wedge osteotomy where shortening of the distance between the tibial tubercle and the tibial plateau occurs, which results in secondary shortening of the patella tendon [2, 8, 21, 22]. This can easily be assessed with preoperative radiographs using the Insall-Salvati ratio , which is the ratio of the patella height to the length of the tibial patella tendon [26]. Patella infera is defined as a ratio of 0.8 or less.

Patella infera is also a problem with respect to elevation of the joint line . The easiest way to compensate for this intraoperatively is to resurface the patella with a smaller than templated patella button placed as proximally as possible. Alternatively, up to 5 mm of extra proximal tibia can be resected, while minimizing the bone resection from the distal femur. This lowers the joint line, or at least insures that the joint line is not elevated, which can improve the patella infera [27]. Finally, at capsular closure , an attempt should be made to advance the medial capsule distally on the lateral capsule, pulling the patella proximally. Patella infera is associated with a decreased arc of motion and potential impingement between the inferior pole of the patella against the anterior flange of the tibial prosthesis. Several studies have shown that the presence of patella infera is not necessarily associated with a less successful outcome of TKA for failed HTO [2, 21] (Fig. 20.6).

Many studies, including reports that show no significant difference between primary TKA and TKA after failed HTO, report less flexion in the post-osteotomy group [1, 3, 6]. Amendola et al. [1] reported an average 14° decrease in flexion in the post-osteotomy group, but believed that this did not compromise the overall functional outcome. Poor preoperative flexion and/or poor intraoperative flexion against gravity after capsular closure warns of this possibility.

A fixed flexion deformity (FFD ) can occur in patients after an HTO. The majority of cases of FFD can be addressed intraoperatively. Care must be taken if the patient has patella infera and a FFD, because the former requires a minimal distal femoral resection to avoid elevating the joint line, while a FFD is often addressed by resecting more distal femur than usual. Careful removal of all posterior osteophytes with the addition of capsular stripping from the femur and tibia can be helpful.

The post-osteotomy joint line is invariably distorted. First, after a closing wedge osteotomy, there is a valgus angulation of the tibia on the coronal view. Second, there is sometimes a loss of the normal posterior slope of the proximal tibial joint line on the sagittal view. In contrast to the anatomical deformity expected with a varus knee, the post-osteotomy valgus angulation of the joint line results in a thicker medial tibial resection than on the lateral side. The tibial cut should resect minimal or no bone from the lateral tibia, with any remaining bony defect managed with lateral augmentation or a structural bone graft if the defect is uncontained. A contained defect can be managed with morsellized graft or cement as required. With preoperative radiographic templating for the appropriate tibial cut, this should be identified hence eliminating intraoperative error (Fig. 20.7).

An osteoarthritis-induced valgus deformity of the knee will be due to a valgus deformity in both the femur and the tibia, whereas a valgus deformity post-osteotomy will be solely due to the tibial deformity. The tibial valgus deformity is compensated for by the varus deformity of the femur due to the initial medial compartment osteoarthritis that necessitated the original HTO. Mont et al. [21] stress the practical implication of this for the surgeon who, after making the routine valgus femoral cut, will make the valgus deformity worse.

The loss of the normal posterior tibial slope can present as either a neutral slope or in fact as an upsloping joint line (Fig. 20.8). The posterior slope must be recreated, necessitating minimal bony resection from the anterior proximal tibia to avoid excess posterior bony resection. Otherwise the potential for flexion and extension gap mismatch can occur, with resultant flexion instability . Once again, radiographic templating will prepare the surgeon for this unusual situation.

A closing wedge osteotomy has no inherent rotational stability other than that provided by the internal fixation. Inadvertent intraoperative tibial rotation or loss of fixation can result in either internal or external rotation of the tibia. As a result the medial one-third of the tibial tubercle may not necessarily be an accurate or reliable guide to tibial rotation. This will necessitate rotation to be determined from more distal landmarks, including the tibialis anterior tendon, the bony ridge of the tibial diaphysis, or the midpoint of the talus. It should be noted that external rotation of the distal tibia increases the Q–angle, which accentuates abnormal patellofemoral mechanics. Difficulty of surgical exposure also produces a tendency to internally rotate the tibial component, which increases the likelihood of patellofemoral subluxation.