Cementing is recommended when the bone defect is smaller than 5 mm in depth regardless of whether it is a contained defect or a peripheral defect. Chen and Krackow demonstrated that cementing is mechanically as strong as metal augmentation in smaller bone defects since the cement fills the defect completely. Cementing is not recommended in relatively younger and active patients. Considering revision surgery, cement is mechanically weaker than bone graft, and bony incorporation occurs easily in younger patients.

When there is a slightly larger peripheral bone defect, cementing and screw augmentation techniques may be used. Ritter and Harty found that cement with screw augmentation showed superior clinical results than those in cement without screw augmentation. Scott introduced screw augmentation in the distal part of the femur and when the appropriate length of screws is used, they can maintain proper alignment as well as strengthen the cement fixation. The technique requires insertion of screws into the bone and covering them with cement. The use of titanium cancellous screws is recommended, and it is necessary to ensure that the screws do not protrude from the cement (Fig. 5.5).

On the other hand, according to Brooks et al., cementing of larger bone defects worsens the clinical results, and radiolucent lines appear beneath the cement. The reason for this is that larger lumps of cement cause thermal necrosis, infection, and fibrous membrane formation. However, Pagnano et al. insisted that osteolysis is the result of poor cementing technique, and they emphasized the importance of proper cementing technique.

It is important not to make a wedge-shaped cut but to make a step cut on the host bone so as to reduce the shear force, especially in a peripheral defect. Chen and Krackow experienced 100 % failure when cement was formed into a wedge shape.

Metal augment is convenient and simple to use. Brand et al. observed good results using metal augments without radiolucent lines after 5–6 years of midterm follow-up.

There are two types of metal augments, wedge or flat block (rectangular) type for the tibial defect and single unit that augments the anterior, distal, and posterior defects simultaneously or separate blocks that augment the femoral defect according to their location.

An autogenous graft, allograft, or both can be used in combination as a bone graft. Once the bone graft is incorporated, it is the most biologic reconstruction. This is the reason why bone graft is recommended for younger and more active patients, even though the bone defect is mild. Dorr et al. stated that the use of a bone graft is indicated when surface defects are more than 50 % and are deeper than 5 mm in one condyle.

A contained defect can be treated with morcellized cancellous graft, but a peripheral defect requires a structural graft. Peripheral defect can be reconstructed using impaction allograft with wire mesh.

Tibial defect may exist on the lateral side in case of a valgus deformity, revision, or infection. However, the most frequent tibial defect in primary arthroplasty is located on the medial side.

In case of a small contained defect, the surface is drilled and the defect is filled with cement or morcellized cancellous graft, and if the defect is large, strut cancellous grafting is performed.

For a peripheral defect, the size of the defect is measured and a solution according to the size of the defect is determined. If the real defect is less than 5 mm, it can be reconstructed by cementing alone or cementing with screw augmentation. If the defect is confined to the epiphysis (real defect less than 10 mm), it can be reconstructed with a metal augment. If the real defect is larger than 10 mm, an allograft is needed. As the osteotomy level from the articular surface is different between medial and lateral condyle, real defect implies the defect needed to be reconstructed. When extension stem is used, the posterior slope should be 0°, since the angle of most of the extension stems is designed perpendicular to the tibial plate.

In case of large cavitary defect, trabecular metal augment can be used (Fig. 5.13).

Special care should be taken for the correct positioning and rotational alignment of the tibial prosthesis as the grafted bone appears different from the normal bone and may adversely affect implant positioning. For positioning and rotational alignment, the tibial tuberosity should be used as the landmark, or the dynamic method can be used for checking the patellar tracking while flexing and extending the knee joint repeatedly. If this is not done, the prosthesis is likely to be inserted in the internal rotation position, and this may cause patellar maltracking or restrict the flexion motion.

For ensuring stability of the bone graft, the use of an extension stem and cementing is recommended in AORI type 3 bone defect, although opinions of surgeons vary in terms of cementing around the stem.

Reconstruction of the femoral bone defect is different from that of the tibial bone defect in several aspects.

First, the anatomy of the femur is complex, and it is often difficult to make an accurate diagnosis of the site and size of the femoral defect on simple X-rays. Second, the joint line should be preserved as much as possible. The tibial defect can be reconstructed up to a certain level, and the gap can be adjusted using different PE thicknesses, whereas in case of the femoral defect, a more accurate reconstruction is required in order to maintain the joint line and to balance flexion and extension gaps. Third, rotational alignment should be determined at the time of bone reconstruction. In almost all cases, the anatomical landmarks are lost due to the bone defect and ordinary osteotomy instruments cannot be used. When determining the rotational axis, the use of the posterior condylar axis is difficult and less reliable due to the bone defect. The external rotation should be based on the transepicondylar axis or Whiteside’s line, but very often it may be impossible to use these parameters. Therefore, it is a good option to set external rotation using the ligament tension technique.

The key to bone defect reconstruction is preservation of the joint line and balance the flexion/extension gap, and the amount of bone defect is measured based on the joint line. The landmarks for determining the joint line include the inferior pole of the patella, fibular styloid process, tibial tuberosity, and femoral epicondyles. The X-rays of contralateral knee joint and the pre-marked joint line before removing the implant in revision surgery (this has been described in Chap.​ 9) can be of great help. The position of the inferior pole of the patella changes during flexion and extension motion which causes the surgeons to make mistakes. The use of fibular styloid process is also difficult as it is located far posteriorly and laterally, while the use of the tibial tuberosity is ambiguous as it is too broad. It is known that the joint line is about 25 mm distal to the lateral epicondyle and 30 mm distal to the medial epicondyle. It is about 20 mm proximal to the tibial tubercle and 14–16 mm distal to the PCL insertion. When the trial implant is inserted at the level of the joint line, the gap between the trial implant and the host bone is the size of the bone defect (Fig. 5.20).

The method of reconstructing the femoral defect is not very different from that of the tibial defect. An F1 defect is filled with autogenous slice bone graft or cement. Resection of the distal portion can be done occasionally according to the gap discrepancy. In F2 defect, filling the bone defect with metal augment is a better option than distal bone resection. However, cementing is not recommended as it can cause malalignment. For an F3 defect, an allograft is required, while tumor prosthesis may also be used in case of a massive bone defect.

Anterior femoral defect is not included in the AORI classification although it is found in almost all cases of revision surgery. It occurs due to stress shielding of the anterior part, or its occurrence is inevitable during implant removal. The defect in this portion can be filled with cement, bone graft, or metal augment. Dennis recommended a cementing technique as the bone graft tends to be resorbed due to stress shielding.

Posterior femoral defect can be treated with metal augment, if the defect is large. If the defect is small, it can be filled with cement. In this case, careful attention should be paid to prevent internal rotation of the femoral component.

For stabilization of the implant and bony incorporation, the use of an extension stem can be considered. Extension stem is useful not only for secure fixation, but also for correct alignment. The length of the stem differs according to the method of bone reconstruction and whether cement fixation or cementless fixation is done. When cementless fixation is used, the length of the stem should be at least 100 mm.

When applying the extension stem, a few things need to be considered. First, most of the femoral extension stems have the angle set at 5° or 7° in coronal plane, and hence the distal osteotomy should be done at this angle. Therefore, it is meaningless to calculate the angle between the anatomical and mechanical axis on femoral A–P X-ray preoperatively. Second, the extension stem should not be too thick. In most of the cases, the portal site of the extension stem in the intercondylar area is broad. If medullary reaming is done according to this width, periprosthetic fracture is likely to occur postoperatively due to endosteal thinning in the diaphysis. Reaming of the medullary canal for IM nailing in fracture treatment does not cause serious problems even though there is endosteal thinning as the IM nail extends beyond the site of reaming. However, the extension stem ends at the site of reaming in the diaphyseal area in TKA and may act as stress riser leading to fracture. Lastly, an offset stem may be needed to prevent anterior angulation due to anterior femoral defect and lower position of extension stem which is a common pitfall during revision TKA.

Patellar defect occurs when there is a bone cyst, in case of severe patellofemoral arthritis due to long-standing patellar maltracking or in case of revision.

The management of patella defect is dictated by the thickness of more than half of the patellar surface after osteotomy. If this thickness is more than 11 mm, then the patellar defect can either be drilled without resurfacing, or the defect can be filled with cement along with resurfacing of patella. If more than half of the patellar surface has a thickness of less than 10 mm after osteotomy, one of the non-resurfacing, biconvex patellae or inset patellae can be chosen (Fig. 5.23). Hanssen stated that pain and function improved after patelloplasty in a huge defect in which bone graft is done, and the graft site is covered with soft tissues. Currently, trabecular metal augmentation has been introduced as a method for reconstruction of a severe bone defect of the patella. Nasser et al. and Ries et al. reported that they could improve function and reduce anterior knee pain by using the trabecular metal augment in patients with severe bone defect (Fig. 5.24). Another method for reconstructing a severe patellar bone defect is the K-wire insertion and cementing technique. Resection patelloplasty by trimming of the remaining bone is performed when there is only a cortical shell of bone remaining. Pavizi et al. observed that this procedure delivers better results than patellectomy or patellar resurfacing. However, Barrack et al. demonstrated that about one-third of the patients experienced anterior knee pain and showed poor satisfaction due to difficulty on stair climbing.

Vince et al. introduced gull-wing sagittal osteotomy and bone grafting of the anterior part of the patella in case only a thin cortical shell of bone was remaining (Fig. 5.25).

Buechel introduced the subsynovial iliac graft by using the synovial envelope for augmenting the bone after resection (Fig. 5.26).

In the coronal plane, 4°–7° valgus of the femorotibial angle is considered normal, and a valgus angle of 3° or lesser is considered as a varus deformity, which is coincident with the mechanical axis of less than 0°.

The principle of treatment varies according to the degree and cause of the deformity. If the deformity is mild, the method of treatment is similar to the usual osteotomy and medial release. In moderate deformity, the semimembranosus or pes anserinus tendon is released carefully. Osteotomy is performed according to the usual method of aligning the mechanical axis. If the varus deformity is severe, it should first be ascertained whether the deformity is a dynamic deformity or static deformity. Dynamic deformity is purely due to a lax LCL and it is obvious while walking. A lateral thrust or lateral instability is a common finding when the varus deformity is caused due to lax lateral structures. This condition can be detected on the preoperative weight bearing or stress X-rays (Fig. 5.27). This type of deformity can be corrected by a medial release and by using a thicker PE. On the other hand, static deformity is accompanied by a bone defect and a contracture of medial soft tissues. This type of varus deformity is mostly caused due to a bone defect of the medial tibial condyle, which leads to secondary contracture of the medial soft tissues. Therefore, ligament balancing should be done simultaneously with the reconstruction of bone defect.

There are a few technical tips for the correction of varus deformity. First, the cartilage of the medial femoral condyle may be worn out. So, it is necessary to verify the alignment from time to time to prevent varus alignment during femoral osteotomy. It is also necessary to increase the amount of external rotation when the varus deformity is severe. Second, balance in extension is critical for the varus deformity, while balance in flexion is critical for the valgus deformity. In other words, imbalance in extension causes more problems as a result of instability due to varus deformity, whereas valgus deformity causes more problems in flexion due to contracture of soft tissues. Third, in patients who have had a severe varus deformity since their young age constitutional varus deformity due to such as Blount disease, less correction of deformity is preferable. Otherwise, the patients may fail to adapt to the corrected neutral alignment and suffer from inconvenience or discomfort. Magnussen et al. analyzed the patients with preoperative varus deformities and reported that residual postoperative varus deformity after TKA is not associated with lower International Knee Society (IKS) scores and increased failure rate, provided the tibial component varus is avoided. Fourth, joint instability is likely to occur when the varus deformity is not fully corrected. If ligament balancing cannot be achieved by any of the methods, a constrained implant should be used.

Lateral tibiofemoral subluxation and rotational deformity may develop after correction of a severe varus deformity. Over-release of the medial soft tissues evokes lateral tightening when PE thickness is adapted to the widened medial gap. In such a case, the insertion of the popliteal tendon can be partially released from its femoral attachment to avoid postoperative painful snapping on the lateral aspect of the knee joint.

A valgus deformity is generally caused due to rheumatoid arthritis or post-traumatic arthritis. Ranawat et al. demonstrated that about 10 % of the patients of total knee joint arthroplasty fall into this category. A valgus deformity is sometimes found in both knee joints, but there also is a condition called “windblown knees” in which one knee has a valgus deformity and the other knee has a varus deformity (Fig. 5.28). The mechanism for windblown knees is that the valgus knee pushes the other knee in a high adduction moment and causes a varus deformity of the other knee joint.

The valgus deformity has more functional disturbance due to more difficulties in walking and provokes more cosmetic problems. Valgus deformity is not found frequently, and hence beginners would face a completely different situation from that of a varus deformity from ligament release to osteotomy.

The common findings of a valgus deformity are as follows: first, the MCL is lax and lateral soft tissue structures are contracted; second, there often is an accompanying lateral bone atrophy or defect usually in the femur; third, it is also accompanied by the patellofemoral problems. Patellar dysplasia, patellar subluxation, and patella alta may also coexist. The knee joint tends to become unstable due to the MCL relaxation and hyperextension.

Therefore, in a valgus deformity, it is required to release the contracted lateral soft tissues, align the knee through bone resection or reconstruction, and correct the patellofemoral malalignment. The degree of deformity is commonly classified as mild if the valgus is less than 10°, moderate if the valgus is less than 20°, and severe if the valgus is greater than 20°, and step by step soft tissue release is recommended according to the degree of deformity.

Insall suggested releasing the iliotibial band at Gerdy’s tubercle and releasing the soft tissues from the lateral condylar area of the femur in a mild deformity, and followed by resecting the iliotibial band and intramuscular septum in a moderate to severe deformity.

Buechel recommends paramedial arthrotomy when there is a valgus deformity less than 8° and lateral arthrotomy when the valgus deformity is greater than 8°. He recommends performing subperiosteal elevation of the ITB from Gerdy’s tubercle in cases with a mild valgus deformity less than 8° as a first step and proceeding to the second step if valgus deformity becomes 0° with a varus stress and returns to the valgus state between 5 and 8° without stress. The second step is subperiosteal elevation of the lateral collateral ligament (LCL) and popliteus tendon, while the proximal portion is attached to the femur. If the valgus deformity cannot be corrected after the second step, he recommends proceeding to the third step to perform subperiosteal elevation of the LCL and biceps femoris complex at the fibular head or resection of the proximal portion of the fibular head.

Krackow classified the valgus deformity into three types. Type I involves lateral femoral bone loss, lateral soft tissue contractures, and intact medial soft tissues. Type II is type I with lengthened medial soft tissues. Type III is severe valgus deformity with malpositioning of the proximal tibial joint line. He suggested advancement of MCL in type II deformity.

It is controversial as to which ligament should be released first and how much release should be performed. Since the iliotibial band (ITB) is the main structure that can cause the deformity, some surgeons emphasize that it should be released first and the popliteus tendon and the LCL should be retained until the very end, if possible. By releasing ITB, primary correction of the deformity is accomplished by eliminating the main cause of deformity, and the possibility of injury to the peroneal nerve due to excessive retraction is reduced. Releasing can be done by performing pie crust technique, a transverse cut, Z-plasty, or periosteal elevation, and the level of release can be about 8–10 cm above the joint line, at the level of the joint line, or at the Gerdy’s tubercle.

On the other hand, others emphasize that the LCL and the popliteus tendon should be released first from the femur based on the theory that releasing the shorter and closer ligaments affects the gap discrepancy between the medial and lateral sides to a lesser extent in flexion, whereas the longer ligaments open up the gap wider in flexion.

In a severe deformity, it is necessary to release the posterolateral capsule as well as the iliotibial band, along with the LCL and the popliteus tendon. If the LCL and popliteus tendon are not released sufficiently, ligament balance cannot be achieved, and this is similar to medial release in which complete release of the MCL is required until ligament balance is established. The LCL should be completely released by cutting the ligament or through periosteal elevation. Other methods include the lateral cruciform retinacular release. It is also possible to perform sliding osteotomy of the lateral femoral condyle or perform fibular head resection.

Some surgeons suggested minimal soft tissue management and using the constrained type, since there are so many complications due to extensive soft tissue release. Easley et al. reported that primary constrained condylar knee arthroplasty in the elderly patients with valgus knee resulted in significant pain relief and improved function.

It is also controversial whether peroneal nerve exploration is needed during soft tissue release. In fact, the reason to release the LCL on the femoral side is to reduce the incidence of peroneal nerve injury as much as possible, hence it is more acceptable that exploration is not necessary unless there is a severe deformity with a flexion contracture.

The osteotomy is performed as usual following the ordinary classic method of aligning with the mechanical axis. And then the valgus deformity is automatically corrected after ligament balancing. However, distal osteotomy of femur should not be done excessively from the beginning as there is atrophy of the lateral femoral condyle. It needs to be done carefully in order to maintain the balanced joint gap. Conservative femoral and tibial resections are performed, when the deformity is combined with recurvatum or medial soft tissue stretching. Thus, soft tissue balancing can be achieved without elevating the joint line or creating a too large extension gap. Scott emphasized that less than 5° of valgus osteotomy should be done to reduce the tension of the medial side even though the difference between the anatomical and mechanical axes exceeds 5°.