Surgery of the Knee

Primary total knee replacement‌‌	279	Distal femoral osteotomy	309
Revision total knee replacement‌	293	Proximal tibial osteotomy	313
Patellofemoral replacement	302	Knee arthrodesis	317
Unicompartmental knee replacement‌	305	Viva questions	320

 

	Range of motion	Position of arthrodesis
Flexion	150°	10°–15°
Extension	0 to −5°
Internal/External rotation	10°	10°

 

11 Surgery of the Knee

Alexander D Liddle, Lee A David and Timothy WR Briggs

 

 

Primary total knee replacement‌

Preoperative planning

Indications

Total knee replacement (TKR) is indicated in the treatment of pain and deformity from the following conditions, when non-operative management has failed or is futile:

• Osteoarthritis

• Post-traumatic osteoarthritis

• Rheumatoid arthritis and other inflammatory arthropathies

• Spontaneous osteonecrosis of the knee (SONK)

   

  Contraindications

• Active or recent local or generalised infection

• Critical arterial ischaemia

• Non-functioning extensor mechanism

• Severe neurological disorders (relative)

• Age (relative): Very young or very old patients should be carefully selected depending on severity of arthritis, level of symptoms and quality of life

  Severe deformity or instability may be a contraindication to the use of an unconstrained, condylar implant and may require the use of a semi-constrained and stabilised or a constrained, hinged prosthesis (see section ‘Revision total knee replacement’, p. 293).

   

  Consent and risks

• Infection: 1%–2% in the general population. Increased in diabetics, smokers, those with a high body mass index (BMI) and those with a history of infection.

• Bleeding: Haematoma formation increases the risk of wound problems, arthrofibrosis

  and infection. Hypovolaemia and anaemia may cause cardiovascular, cerebral or renal complications.

• Venous thromboembolism: Below-knee deep vein thrombosis (DVT) occurs in

  approximately two-thirds of patients following TKR. The risk of fatal pulmonary embolism (PE) is approximately 0.1%. The prevention of DVT and PE remains a controversial topic, but it is almost universally accepted that mechanical and some form of chemical thromboprophylaxis should be used.

• Neurovascular injury: Damage to the infrapatellar branch of the saphenous nerve during

  the incision is often unavoidable and leads to sensory change on the anterolateral shin; this occurs in upwards of 70% of patients but generally improves over time. Damage to important nerves and blood vessels is rarer, and can be caused by direct transection, traction or pressure. Discrete arterial damage is rare (approximately 0.05%) but must be recognised and dealt with immediately. Distal arterial thromboembolism must be promptly recognised, pressure dressings released and a vascular surgical opinion sought. Common peroneal nerve (CPN) injury has an incidence of approximately 0.5% and should initially be managed by release of pressure dressings with exploration indicated if caused by haematoma. Traction injury to the CPN is most common if a severe valgus deformity is corrected, particularly if combined with a fixed flexion deformity. In these cases, a foot drop splint should be used to prevent equinus contracture and nerve conduction studies may be performed at a later date.

• Fractures: The risk of fracture is increased in osteoporosis and rheumatoid arthritis.

  Significant notching of the anterior distal femoral cortex is thought to increase the risk of postoperative periprosthetic fracture, but the evidence is weak. Excessive patella resection during resurfacing increases the risk of patella fracture. Intraoperative fractures usually require immediate fixation and the use of stemmed implants.

• Extensor mechanism injury: Avulsion of the patellar tendon is a disastrous complication

  and must be avoided as it severely compromises the outcome following TKR. In the event of this occurring, the tendon must be reattached to the tibial tuberosity and protected, although the result is usually poor.

• Stiffness may be caused by true arthrofibrosis, but other causes must be ruled out. These

  include infection and mechanical problems, such as oversizing the femoral component, errors of rotation leading to patellar maltracking, reversing the tibial slope or making an inadequate bone resection. Treatment depends on the underlying problem.

• Instability may be caused by unequal flexion/extension gaps, soft tissue imbalance,

  ligamentous insufficiency, insufficient insert thickness, polyethylene wear or patellofemoral maltracking. Treatment depends on cause.

• Revision surgery: 4% at 10 years according to worldwide national joint registries. A recent

meta-analysis suggested that over 80% of TKRs survive to at least 25 years. The majority of patients undergoing knee replacement will never undergo revision.

 

 

Operative planning

Clinical examination should pay careful attention to alignment, deformity, instability, range of movement and extensor mechanism function. Scars should be carefully noted, and a distal neurovascular assessment must be performed.

Recent weightbearing anteroposterior, lateral and skyline radiographs must be available and long-leg alignment views are helpful to establish the mechanical axis of the leg (Figure 11.1). It is imperative that the patient’s symptoms should correlate with the radiographic findings. Templating of preoperative radiographs should be performed if possible, and it is the responsibility of the surgeon to ensure that the required implants are available.

 

 

us

 

 

Figure 11.1 The mechanical and tibiofemoral axes of the lower limb.

 

Choice of implant

Most implants in current use are evolutions of the condylar knee designs popularised in the late 1960s and early 1970s; the great majority are cemented. In most cases, the femoral component is polyradial in the sagittal plane to recreate the ‘J-curve’ of the native

 

condyles. The tibial component is usually modular with a metal baseplate and an ultra-high molecular weight polyethylene insert, although monoblock (metal-backed or all-polyethylene) tibial components have good published results and are generally cheaper. The tibiofemoral articulation is minimally constrained, but the tibial component is dished to mitigate for the loss of the anterior cruciate ligament. In cruciate retaining (CR) implants the posterior cruciate ligament (PCL) is retained; in posterior stabilised (PS) designs, the PCL is resected and a cam-post mechanism is used to maintain stability; in most studies the outcomes are similar. As designs have evolved, the femoral components have gone from being symmetrical in the coronal plane to being sided (with the asymmetry either restricted to the trochlear flange or involving different radii of the medial and lateral condyles as in the native knee); newer designs are generally more conforming to prevent subtle but symptomatic instability. Variations include rotating platform tibial components, medial rotation knees where the medial side represents a ball and socket joint and the lateral side is unconstrained, and femoral components with a single radius of curvature (which is designed to maintain collateral ligament tension through the range of flexion). There is little published evidence to suggest the superiority of one philosophy over any other.

In the end, the choice of implant remains at the discretion of the surgeon or, increasingly, the purchasing agreements of the trust. It is the surgeon’s responsibility to ensure that he or she is familiar with the implants used, that all necessary equipment is available and that the appropriate range of sizes are readily to hand.

 

Anaesthesia and positioning

Anaesthesia is usually general, regional or combined, depending on the preferences of the anaesthetist and surgeon and the patient’s co-morbidities.

The patient is positioned supine on the operating table with a lateral thigh support and foot bolster, allowing free flexion and extension of the knee. Pressure areas should be protected with gel pads. Most, but not all, surgeons use a tourniquet for all or part of the procedure unless contraindications exist (such as arterial insufficiency). The tourniquet should be well padded and placed high on the thigh. Tourniquet time should be clearly documented and should not exceed 2 hours. A dose of an appropriate antibiotic is administered intravenously prior to the inflation of the tourniquet. The skin in the area of the incision should be shaved immediately prior to surgery. The surgical field is prepared with an antiseptic solution. The foot should either be thoroughly prepared or wrapped with an impervious ‘shut-off’ drape. Appropriate waterproof drapes should be carefully applied. An antibacterial, transparent adhesive drape is usually applied to the surgical field.

 

Surgical technique

By far the most common approach to the knee joint in TKR is the medial parapatellar approach, which is discussed later. The subvastus, midvastus and direct lateral approaches are used much less frequently. Other extensile approaches are discussed in the section ‘Revision total knee replacement’.

 

Landmarks and incision

The position of the patella, patellar tendon and tibial tubercle should all be noted. An anterior midline longitudinal incision is made, usually with the knee in flexion. The incision needs to be long enough to allow adequate exposure and avoid excessive skin stretching; this runs proximally from the level of the tibial tubercle for approximately 20 cm, although the length is heavily dependent on the patient’s build.

 

Dissection

 

Structures at risk

The medial collateral ligament (MCL) may be damaged during medial release. The risk of this can be minimised by careful subperiosteal release using either a periosteal elevator or coagulating diathermy.

The patellar tendon may be damaged during excision of the fat pad, which can be prevented by always cutting away from the tendon itself. The patellar tendon may be avulsed at its insertion to the tibial tubercle during eversion of the patella and flexion of the knee. This is a disastrous complication and can be prevented by extending the deep dissection proximally, dividing any lateral plicae and performing a lateral parapatellar release to allow eversion of the patella. External rotation of the tibia also relaxes the extensor mechanism.

 

 

Dissection continues in the midline, until the quadriceps tendon is identified. The medial and lateral skin, subcutaneous fat and deep fascia should be reflected in matching thick flaps to allow exposure of the quadriceps tendon, medial patellar retinaculum and patellar tendon.

The medial parapatellar incision is extended from the quadriceps tendon proximally, through the medial parapatellar retinaculum and along the medial border of the patellar tendon distally (Figure 11.2). There should be at least a 3 mm cuff of quadriceps tendon left attached to vastus medialis and a cuff of medial retinaculum attached to the patella to allow closure.

The medial capsule is released subperiosteally off the proximal tibia to gain exposure to the medial compartment. In a varus knee, this dissection should include the deep medial collateral ligament and, depending on the degree and correctability of the varus deformity, may extend as far as the posteromedial corner. In a valgus knee, this medial release should be kept to the minimum required to allow exposure.

With the knee in extension, the patella is everted and the knee flexed. The retropatellar fat pad may be partially or fully excised if necessary. The visible remnants of the medial and lateral menisci may be resected at this stage and the anterior cruciate ligament (ACL) must be divided and resected. If a PS implant is to be used the PCL can be resected now by dissecting it from its femoral attachment with diathermy. Osteophytes may be debrided at this stage.

 

 

 

 

 

Figure 11.2 The medial parapatellar approach to the knee.

 

Procedure

The primary goals of surgery are relief of pain, restoration of function and longevity of the prosthesis. The immediate technical aims of the operation are to restore alignment, allow a good range of motion, good stability and ligamentous balance throughout the range of motion with good patellar tracking.

The primary requirement to allow a TKR to function is for the gaps between the femur and the tibia to be symmetrical and equal in flexion and extension. This can be achieved by one of two main routes, ‘measured resection’ and ‘gap balancing’. Measured resection is the more common approach and involves performing anatomically consistent bone cuts on the femur and tibia and restoring symmetry and equality of gaps by performing sequential soft tissue releases. In gap balancing, the femoral and tibial cuts are made in order to maintain the tension of the medial and lateral collateral ligaments without significant releases. Some surgeons will perform a measured resection of the distal femur and proximal tibia with flexion gap and femoral rotation achieved using a gap balancing technique. Either way, the surgeon must appreciate that a TKR is as much a soft tissue operation as a bony procedure.

 

Bone cuts

In most cases, the bony cuts are made using a measured resection technique using standard instrumentation. Whether the femoral or tibial cut is made first depends on the surgeon’s preference and type of prosthesis used; however, it is advisable to perform the distal femoral and proximal tibial cuts before final femoral preparation. This ensures that the extension gap is adequate and symmetrical, and allows the flexion gap (dictated by the anteroposterior size and position of the femoral component) to be fine-tuned to match the extension gap.

 

 

Structures at risk

• The MCL must be carefully protected during saw cuts.

• Patellar tendon.

• Common peroneal nerve may be injured by injudicious placement of the lateral retractor or by stretch in large corrections.

• Popliteus tendon can be damaged during posterior femoral saw cut or resection of meniscus.

• Popliteal vessels and tibial nerve can be injured during removal of posterior osteophytes, posterior capsular release, PCL resection and when cutting the posterior tibial cortex with the saw, if not protected. The anatomy of the popliteal artery in relation to the knee joint is extremely variable.

 

 

Femoral cuts

The femur should be prepared with the use of an intramedullary alignment jig if possible. The tibial cut can be made by using intra- or extramedullary alignment jigs, depending on the surgeon’s preferred method and the degree of extra-articular deformity of the tibia. There is evidence to show that intramedullary referencing of the tibial cuts is more accurate, but it has also been shown to increase the risk of fat embolism.

Femoral preparation is undertaken with the knee flexed and the patella everted. A large drill bit is used to create an entry point in the distal femoral canal at a point approximately 1 cm anterior to the insertion of the PCL within the trochlear notch. The intramedullary rod should be inserted into the canal with care, especially if a previous total hip replacement has been performed. The distal femoral cutting jig is positioned over the rod and adjusted so that the distal cut is set at a 5°–9° valgus angle to the appropriate side of the knee to be replaced (Figure 11.3). Ideally, this should be chosen to match the anatomical axis of the contralateral limb, if normal.

The distal cutting jig is secured with two or three pins that should be fully inserted to ensure that the saw is not hampered and to allow the saw blade to make ample excursion to complete the cut. The amount of distal femoral resection performed depends upon the thickness of the implant (usually around 9 mm) and any fixed flexion deformity present (see later). It is imperative that the medial and lateral soft tissues are retracted and protected with either Hohmann or Trethowan retractors. The cut bone surface should be of sufficient surface area and quality to allow adequate fixation and must expose trabecular bone. If the distal femur is particularly sclerotic in parts, a ‘second pass’ with the saw blade may be required to achieve a flat surface, but one must bear in mind that repeated passes with a power saw generates heat, necrosis and metal debris from the jig.

The distal femur must then be sized to enable placement of the appropriate cutting block. Sizing jigs generally work on an anterior or posterior referencing system, using either the anterior distal femoral cortex or the posterior femoral condyles as the baseline, measuring the amount of anteroposterior resection required accordingly. If the sizing is perfect, the size and position of the implant will be the same whether anterior or posterior referencing is used. If there is over- or under-sizing, then the use of a posterior referencing system will

 

 

 

 

Figure 11.3 Distal femoral resection.

 

lead either to notching of the anterior femur or ‘overstuffing’ of the patellofemoral joint. If an anterior referencing system is used, then a gap mismatch will occur, with the flexion gap being either tight, resulting in stiffness, or loose, resulting in instability in flexion. If the knee is found to be between sizes of implant, generally the smaller implant is chosen. The typical sizing jig has an anterior stylus that must be seated down onto the anterior cortex, and it may be necessary to remove the overlying synovium to ensure accurate sizing. When the desired size is estimated, marker holes are made on the distal femur through the appropriate holes on the jig, to enable positioning of the femoral cutting block. Attention must be paid to femoral rotation as mistakes (usually too little external rotation) will result in flexion gap asymmetry and patellar maltracking. Rotation should be set parallel to the interepicondylar axis, which is perpendicular to Whiteside’s line. The transepicondylar axis is usually around 3° externally rotated to the posterior femoral condyles and in most systems, the holes made by the femoral sizing jig are set in 3° of external rotation to the posterior condylar axis to match this. In the valgus knee the lateral condyle may be hypoplastic, and relying on the jigs may result in internal rotation of the femoral component. In such cases, additional external rotation should be introduced, guided by the transepicondylar axis.

The cutting block corresponding to the measured size is placed onto the cut surface of the distal femur, with pegs sitting into the previously drilled marker holes. Cuts (particularly the anterior cut) can be estimated using an ‘angel wing’. The cutting block is firmly impacted until seated flat onto the cut surface of the distal femur and secured with obliquely placed pins. Again, the soft tissues must be carefully retracted during the placement of instrumentation. If there is any difficulty in seating either the sizing jig or cutting block, the surgeon must check that all osteophytes are removed, that there is adequate meniscal

 

resection, that the bone cuts are complete and that the soft tissues are retracted sufficiently. Anterior and posterior cuts should be made prior to the chamfer cuts to prevent destabilising the cutting block. If it is apparent that there will be significant notching of the distal femur, the cutting block should be removed and the sizing reassessed (Figure 11.4). If there is a possibility of minor notching occurring, this should be controlled and any sharp edge of anterior cortex should be smoothed off with the saw or a bone file. The cut bone fragments can then be removed with knife and forceps and the posterior condylar cuts can be removed with a broad osteotome. The distal femur is then examined to ensure that the cuts are complete. Large posterior osteophytes apparent on the preoperative lateral radiograph or evident after bone cuts can be removed by lifting up the femur and carefully using a broad osteotome under direct vision.

 

 

1

2

3

 

 

 

Figure 11.4 The anterior femoral cut.

 

Tibial cut

The tibial cut should be made perpendicular to the axis of the tibia in the coronal plane ensuring a sufficient anterior to posterior slope (Figure 11.5). If intramedullary referencing is used, the entry point should be made with a drill at the centre point of the tibia. The intramedullary rod should be inserted comfortably into the tibial canal. With intramedullary referencing, the slope is generated using an appropriately angled cutting block. If extramedullary referencing is used, the rod should be in line with the tibial tubercle, and the distal tip of the rod should lie just medial to the centre of the ankle joint (as this is where the mechanical axis of the limb passes). Use of anatomical landmarks in the foot, such as the second metatarsal, is less reliable as rotation can occur within the hindfoot and midfoot. With extramedullary referencing, the anteroposterior slope of the tibial cut can be introduced either by use of an angled cutting block (as with the intramedullary technique) or by adjustment of the extramedullary jig itself. Depth of resection is estimated using a stylus. Generally, either the depth of the implant (usually 8–10 mm) is taken from the preserved tibial plateau (the lateral plateau in varus arthritis) or 2 mm is taken from the depth of the wear scar to ensure an adequate resection has been made. In valgus disease, the preserved tibial plateau is the medial, which is concave, and it may be appropriate to take a smaller resection than would be appropriate in varus disease.

 

Figure 11.5 The tibial cut.

 

After the tibial resection is complete, the remaining meniscal remnants can be excised and the tibial component is sized. Following a trial of the components, the tibia can be prepared to accept the stem or keel of the prosthesis. Sizing the tibial component is a balance between achieving adequate coverage of the cut surface to provide support to the implant and avoiding overhang which may cause symptoms. The midpoint of the tibial component should be in line with the medial third of the tibial tubercle, but the surgeon should err on the side of external rotation. Often, particularly with symmetrical tibial components, an oversized tibial component may fit the cut surface but only with the introduction of excessive internal rotation. In this situation, a smaller size should be chosen and care taken to position the implant accurately.

 

Balancing the knee

Balancing flexion/extension gaps

As stated previously, prior to implantation the flexion and extension gaps should be equal and symmetrical in full extension and 90° of flexion (Figure 11.6). Inequality or asymmetry can be addressed using bony or soft tissue adjustment. Adjustment of the bone resection on the tibia will affect the flexion and extension gaps equally. Adjustment of the bone resection on the femur will affect either the extension gap (adjustment of the distal cut) or the flexion gap (adjustment of the posterior femoral cut) individually.

Figure 11.6 Flexion and extension gaps.

 

Unequal but symmetrical gaps

If the flexion gap and extension gap are different, but the medial and lateral tensions are similar, the following steps can be taken:

 

• Tight in extension, flexion satisfactory

  • Solution: Increase distal femoral resection. Recess the PCL or convert to PS design. Beware of raising the joint line with excessive distal femoral resection.

• Tight in flexion, extension satisfactory

  • Solution: Resect more posterior femur by downsizing the femur (in anterior referencing implant) or translating the femur more anteriorly. Check that the tibial slope is adequate and increase it if necessary.

• Tight in flexion and extension

  • Solution: Increase tibial resection. This may necessitate a smaller size of tibial component due to the tapering morphology of the proximal tibial metaphysis.

• Loose in flexion and extension

  • Solution: Increase thickness of insert. Ensure that laxity has a firm endpoint and no ligamentous injury has occurred. If there is ligamentous injury, a more constrained implant may be necessary (see the section ‘Revision total knee replacement’).

     

    Sofi tissue balancing

    When a measured resection technique is used, soft tissue balancing is used to ensure symmetry of flexion and extension gaps. The extent and sequence of soft tissue releases depends upon the pre-intervention deformity. Often, only minimal releases are necessary –it is better to start with a small release and increase them later than to perform large releases initially which turn out to be excessive.

     

• Varus deformity: This is the most common deformity seen in osteoarthritis and generally leads to a tight medial collateral ligament in extension. The flexion gap may be equal or tight in flexion. The sequence of releases is

  • Resection of osteophytes and subperiosteal release of capsule and deep MCL.

  • Distal extension of release to involve superficial MCL and pes anserinus.

  • Recession or resection of PCL.

  • Release of posteromedial capsule or semimembranosus.

• Valgus deformity: This is the most common deformity in rheumatoid arthritis and can be seen in osteoarthritis and post-traumatic or post-meniscectomy arthritis. The lateral side is tight in extension (commonly), flexion (less commonly) or both. In all cases, a minimal medial release is performed. A PS implant may be chosen. The sequence of releases is

  • Removal of osteophytes.

  • If tight in extension, ‘pie-crusting’ of the iliotibial band or release from Gerdy’s tubercle.

  • If tight in flexion, popliteus release.

  • If tight in both flexion and extension, the lateral collateral ligament may need to be ‘pie-crusted’ or released from the femur.

  • Finally, in a severe combined valgus and fixed flexion deformity, posterolateral capsule and lateral head of gastrocnemius may be released.

• Fixed flexion deformity: This can be the result of posterior osteophytes and contracture

  of the posterior capsule in severe disease. Either bony or soft tissue releases may be performed:

  • Taking care not to excessively elevate the joint line, extra bone may be removed from the distal femur (1 mm for every 2°–3° of fixed flexion).

  • Osteophytes should be removed from the posterior femur with osteotomes and Kocker’s forceps under direct vision.

  • Posterior capsule may be released using a curved osteotome.

  • PCL may be recessed or resected.

 

The patella

Whether the patella should be resurfaced is still a controversial issue. Some surgeons always resurface, some never resurface and some do only if there are patellofemoral symptoms or if the retropatellar surface is severely affected. The evidence on patellar resurfacing is mixed. The Knee Arthroplasty Trial, the largest randomised trial of TKR in osteoarthritis, suggests that patellar resurfacing results in no difference in patient-reported outcome but reduces the rate of revision by eliminating secondary patellar resurfacing. Inflammatory arthropathy is considered to be a disease of the whole joint and patellar resurfacing is

 

mandatory in rheumatoid and other inflammatory arthropathies (unless the patellar bone stock precludes it).

The majority of patella buttons used are monoblock and all-polyethylene. There are two types of patella button – onlay and inlay. As the respective names suggest, these designs utilise a prosthesis that is implanted either onto or into the resected retropatellar surface. To resurface the patella, the knee should be extended and the patella fully everted. Peripheral osteophytes can be removed with a bone nibbler to demarcate the actual articular surface. The thickness of the patella should be measured and the amount of bone/cartilage removed should approximately correspond to the thickness of the implant, although if there is severe damage it may be less. Resecting too little bone runs the risk of overstuffing the knee and if a sclerotic surface is left behind, fixation is compromised, whereas resecting too much increases the risk of fracture. Many implants now have calibrated clamps and jigs that help indicate the correct resection level.

To avoid an increase in the ‘Q’-angle and therefore reduce the likelihood of maltracking, the patella button should be slightly medialised and both the femoral and tibial components should be lateralised and externally rotated. Patellar tracking should be adequate even before the patellar retinaculum is closed. Maltracking of the patella is usually related to problems with rotation of the femoral or tibial components; rarely, if the patella maltracks with well-positioned implants then a lateral retinacular release may be necessary. This should be performed with a diathermy (as it may lead to significant bleeding and bruising postoperatively) and should be carried out from distal to proximal and from deep to superficial. There are often palpable fibrous bands, and release of these is sometimes enough. To enable the release to be performed, the patella is lifted anterolaterally with the knee in extension. If possible, the superior lateral geniculate artery should be preserved to avoid devascularisation of the patella. Superficial releasing with a resultant subcutaneous flap and undermining of the lateral skin should be avoided. Where TKR is performed following a previous patellectomy, a PCL substituting implant should be used in order to avoid excessive anterior subluxation of the femur on the tibia due to an already relatively attenuated extensor mechanism.

 

Implantation of prostheses

While cementless and hybrid designs are available, most TKRs are implanted using cement. In cemented knee replacement, two mixes of antibiotic-impregnated polymethylmethacrylate (PMMA) bone cement should be used. The surgeon should be familiar with the biomechanical properties of the cement and its mixing technique. Following satisfactory trials, the selected components are checked by the surgeon and opened. The knee is flexed and the patella everted allowing the tibia to be subluxed anteriorly, with a Hohmann retractor or similar, and the prepared surface of the tibia exposed medially and laterally with spiked retractors. The knee is washed out thoroughly with normal saline pulsed lavage in order to expose the bone trabeculae and maximise the mechanical fixation of the cement. If sclerotic bone surfaces are present, a small drill can be used to make multiple small ‘key holes’. The knee should be thoroughly dried with suction and swabs. The cement can then be mixed and the whole surgical team should change the outer layer of gloves. In most situations, cementing of both

 

components can be performed simultaneously, but on occasions it may be desirable to perform cementing of the components separately with different mixes of cement.

To ensure a satisfactory and efficient cementation process, everything should be prepared and ordered in a logical fashion. The tibial component is usually implanted first. Cement can be applied onto the surface of the tibia, or the implant, or both, using a gun with short nozzle or a spatula. The tibial component is positioned in the correct orientation and firmly seated with a soft impactor and hammer. Excess cement is removed. Cement is applied to the cut surface of the femur and the posterior surface of the implant taking care not to apply too much cement to the posterior condyles as removal of excess cement can be difficult from the posterior part of the femur. The femoral component must be positioned carefully in relation to the distal femur; in particular flexion of the femoral component should be avoided. The femoral component must be firmly impacted and any excess cement should be removed. Either a trial or definitive insert is attached to the tibial baseplate, the knee is extended and axial compression applied. (Note: Hyperextension leads to uneven cement pressurisation and may cause posterior ‘lift-off’ of the tibial baseplate.) If the patella is resurfaced the orientation should be checked; once positioned, the patella is compressed and held with a clamp. The knee can then be flexed again and any further cement extruded can be removed quickly. The knee is then extended and further axial compression applied.

Closure

Once the cement has set, the knee can be washed out again with pulsed lavage. Some surgeons prefer to deflate the tourniquet and gain haemostasis prior to closure. Intravenous or topical tranexamic acid may be given at this stage. Alternatively, the knee can be closed and a pressure dressing applied prior to deflation of the tourniquet. The use of drains has declined over recent years and the evidence for their use is weak.

The actual closure technique varies with surgical preference, but it is important that the repair is watertight and that range of motion is maintained with no patella maltracking. Closure of the knee in flexion ensures that the correct tension is achieved. The deep layer is closed with a heavy suture (e.g. number 1 Vicryl), by means of a continuous repair of the quadriceps tendon, interrupted repair of the parapatellar retinaculum and continuous repair of the medial capsule to patellar tendon. The deep fascia can be closed as a separate layer if desired or the subcutaneous fat can be opposed with deep interrupted sutures. The deep dermal layer is closed with a continuous absorbable suture to allow tension-free closure of the skin with surgical staples or a continuous absorbable subcuticular suture. A sterile occlusive dressing and a padded compression bandage are applied.

Postoperative care and instructions

Regular neurovascular, cardiovascular and respiratory observations are mandatory. Urine output, temperature and drainage (if a drain is used) should also be monitored. Adequate analgesia should be administered. Mechanical and chemical thromboprophylaxis should be given according to local and national protocols. Haemoglobin levels should be checked 24–48 hours after the procedure. Any drains, urinary catheters, epidural

 

lines and intravenous cannulae should be removed as soon as appropriate to avoid unnecessary portals of infection. Pressure dressings should be reduced and ice applied. Full weightbearing and active range-of-motion exercises should be commenced as soon as possible. The wound should be inspected and radiographs performed prior to discharge should be checked. The patient must be declared safe for discharge and for routine cases should be able to straight leg raise and flex the knee from 0° to 90°.

Skin clips should be removed 10–14 days after surgery, and an outpatient appointment should be arranged approximately 6 weeks postoperatively. Ideally, patients undergoing TKR should be followed up for life with serial radiographs, but in reality this is rarely possible.

 

Recommended references

Bayliss LE, Culliford D, Monk AP et al. The effect of patient age at intervention on risk of implant revision after total replacement of the hip or knee: A population-based cohort study. Lancet. 2017;389: 1424–1430.

Evans JT, Walker RW, Evans JP, Blom AW, Sayers A, Whitehouse MR. How long does a knee replacement last? A systematic review and meta-analysis of case series and national registry reports with more than 15 years of follow-up. Lancet. 2019;393:655–663.

Murray DW, MacLennan GS, Breeman S et al. A randomised controlled trial of the clinical effectiveness and cost-effectiveness of different knee prostheses: The Knee Arthroplasty Trial (KAT). Health Technol Assess. 2014;18;1–235.

Shetty AA, Tindall A, Ting P, Heatley FW. The evolution of total knee arthroplasty. Part III: Surface replacement. Curr Orthop. 2003;17:478–481.

Whiteside LA. Ligament Balancing in Total Knee Arthroplasty: An Instructional Manual. Berlin, Germany: Springer, 2004.

 

Revision total knee replacement‌

This section is not intended as a comprehensive guide to revision knee replacement but rather covers the principles of the procedure. This section refers extensively to the section ‘Primary total knee replacement’ (p. 279).

 

Preoperative planning

Indications

Revision TKR is indicated in the treatment of pain, stiffness or instability from a failed TKR. The cause of failure must be diagnosed prior to embarking on revision surgery: revision for unexplained pain has poor outcomes. Knees may fail for a single reason or several in combination. The most common indications for revision are

• Infection

• Aseptic loosening/osteolysis

• Polyethylene wear

• Instability

• Stiffness

• Patellofemoral dysfunction

• Periprosthetic fracture

   

  Contraindications

• Medically unfit for surgery or anaesthetic

• Critical arterial ischaemia

• Non-functioning extensor mechanism

• Unexplained pain

• Insufficient skin coverage (relative)

• Severe neurological disorders (relative)

• Age (relative): Very elderly patients should be carefully selected depending on severity of symptoms, quality of life and options available

   

  Consent and risks

  All of the risks and complications of primary TKR occur at increased rates following revision. The overall complication rate for revision knee replacement is approximately 25%, while the outcome of a successful revision is significantly inferior to the results of a successful primary. In revision for infection, the best centres report rates of clearance of infection at up to 95%.

  • Infection (or failure to eradicate)

  • Bleeding

  • Venous thromboembolism

  • Wound problems

  • Neurovascular injury

  • Fractures

  • Extensor mechanism injury

  • Stiffness

  • Instability

  • Wear

  • Loosening

  • Pain

   

   

  Operative planning

  A thorough history and examination is essential to rule out pain referred to the knee from elsewhere and to assess the level of pain and functional disability. Special consideration should be given to potential risk factors, and realistic goals should be identified. The examination should pay careful attention to ligamentous instability, range of movement and extensor mechanism function. In cases of apparent aseptic loosening, stiffness or pain, infection must be excluded: inflammatory markers should be performed, and there should be a low threshold for performing an aspiration or biopsy. Scars should be carefully noted, and a distal neurovascular assessment must be performed. If the skin over the knee is of poor quality, it may be necessary to consult a plastic surgeon.

   

  Recent weightbearing anteroposterior, lateral and skyline radiographs must be available, and long-leg alignment views are helpful in diagnosing malalignment. Computed tomography is helpful to assess the degree of bone loss in cases of osteolysis and can be useful in diagnosing rotational malalignment. It is absolutely essential that a cause for the failure is found. Templating of preoperative radiographs should be performed if possible; it is the responsibility of the surgeon to ensure that the required implants are available.

   

  Choice of implant

  When selecting an implant for revision TKR, the aim is to confer sufficient constraint at the joint to address any ligamentous deficiency, to provide sufficient fixation to support the more constrained implant in compromised bone stock, and to fill any bony defects that may be present.

  The native knee joint is effectively unconstrained at its bony surfaces. The femur and tibia are connected by the cruciate and collateral ligaments, capsule, extensor mechanism, hamstrings and gastrocnemius. Unicompartmental knee replacement is similarly unconstrained as all of these structures are preserved. Primary TKR is minimally constrained, introducing a dish to substitute for the ACL and, in PS designs, a cam and post. As more capsular and ligamentous structures are lost, more constraint is necessary – this is the ‘ladder of constraint’ (see box). Generally, except in exceptional circumstances (such as revision of unicompartmental or patellofemoral replacement), a condylar constrained design is the minimum acceptable degree of constraint.

  The more constraint that is present at the knee joint, the more forces are transferred to the bone-implant interface. As a result, increasing constraint necessitates more secure fixation within bone. A useful concept for fixation of revision knee prostheses is that of zonal fixation. Fixation can be achieved in three zones – the epiphysis/joint surface (zone 1), the metaphysis (zone 2) and the diaphysis (zone 3); fixation in two of the three is considered to be adequate. Traditionally, revision implants achieve fixation in zones 1 and 3: sufficient epiphyseal bone is identified by re-cutting, and may be supplemented using wedges or augments (augments are very frequently used on the femoral side to avoid raising the joint line). Zone 3 fixation is achieved using stems, which may be broad to allow a cementless press-fit, or may be narrower and cemented. Fixation in zone 2 can be achieved using metaphyseal sleeves or cones. Sleeves attach to the component via a morse taper to form a monolithic structure. They are wedge shaped and are designed to provide metaphyseal fixation in a similar way to a cementless total hip replacement. If a stem and a sleeve is present, secure fixation in zone 1 is unnecessary. Cones are separate from the implant and come in a variety of shapes – cones are designed to fill defects and provide supplementary fixation in the metaphysis.

  Bone loss can be addressed using bone graft (for contained defects in the metaphysis, for example) or by the use of augments, cones and sleeves. Massive bone loss may necessitate the use of a distal femoral or, rarely, a proximal tibial replacement.

   

   

  Ladder of constraint (from least to most constrained)

• Unicompartmental knee replacement

  • Round on flat or mobile bearing design results in a completely unconstrained joint surface

  • Requires functionally intact ACL, PCL, collateral ligaments

• Cruciate retaining TKR

  • Minimal constraint through dishing at joint surface to substitute for absent ACL

  • Requires functionally intact PCL and collaterals

• Posterior stabilised TKR

  • Additional constraint through cam-post mechanism to substitute for absent PCL

  • Provides no mediolateral constraint so requires functionally intact collaterals

• Condylar constrained revision knee replacement (‘high post posterior stabilised’)

  • Large central cam-post substitutes for PCL and provides a degree of mediolateral constraint

  • Will allow for a degree of ligamentous laxity but not for a completely absent MCL

• Rotating hinge

  • Direct connection of femur to tibia removes the need for any functional collateral ligaments

  • Rotating hinge design reduces the degree of torsion transmitted to interface

  • Requires an intact extensor mechanism

• Fixed hinge

  • Rarely necessary; has significantly inferior survival compared to rotating hinge

  • Useful in those with neuromuscular disorders

  • Can be used to treat patellofemoral dysfunction following rotating hinge

  • Requires an intact extensor mechanism

 

Anaesthesia and positioning

See ‘Primary total knee replacement’ (p. 279). The operation is likely to last longer than a primary knee replacement, leading to more physiological disturbance. It can be helpful to exsanguinate the leg after preparation and draping to save tourniquet time.

 

Surgical technique

Although many revision procedures can be performed via the medial parapatellar approach, as described in primary TKR, other extensile approaches may be required to gain adequate exposure.

 

Landmarks and incision

All scars should be marked with a sterile pen. If possible, a generous midline incision is used. If there are multiple longitudinal incisions in front of the knee, the most lateral scar should be used to avoid necrosis of the intervening strip of skin due to the fact that the blood supply passes from medial to lateral. The incision needs to be long enough to allow adequate exposure and avoid excessive skin stretching.

 

Superficial dissection

Skin flaps should be kept as thick as possible and should not be undermined. The quadriceps and patellar tendons should be defined. Identification of the correct tissue plane is easier if the incision is extended to an area previously untouched.

 

Deep dissection

 

Structures at risk

• The medial collateral ligament is at risk from aggressive synovectomy and medial release.

• The patellar tendon is usually thickened, tight and at risk of avulsion. The patellar

  tendon and quadriceps tendon should be thinned down by excision of any thickened fibrous tissue and the articulating surface of the patella should be exposed. If the patella does not evert or subluxate easily, one or more of the following measures needs to be performed.

• All other important structures around the knee are at greater risk of injury during

revision surgery than in the primary procedure due to scar tissue, difficulty in exposure and stiffness or laxity.

 

 

The standard medial parapatellar approach is usually performed initially. It is usually necessary to perform an extensive synovectomy in order to improve exposure and to recreate the suprapatellar pouch and medial and lateral gutters. The fat pad is excised. Medial release should be performed to allow exposure of the tibia. There is usually a plane visible between the pseudocapsule and normal tissue, and this can be developed with knife or diathermy and the pseudocapsule carefully pulled away under tension. The PCL is usually sacrificed, this can be performed following implant removal.

 

Lateral parapatellar release

It is almost always necessary to perform some degree of lateral parapatellar release to allow eversion of the patella. It is usually beneficial to perform the lateral release early on. The release should be performed from deep to superficial and from distal to proximal, alongside the lateral border of the patellar tendon and lateral retinaculum. To reduce subsequent blood loss, it can be performed using diathermy. Full-thickness lateral release should be avoided if possible, but if this is necessary to gain exposure the superior lateral geniculate artery should be left intact and the lateral parapatellar retinaculum should be closed later.

 

Quadriceps snip

This involves a lateral incision into the quadriceps tendon from the proximal extent of the standard medial parapatellar approach (Figure 11.7a). A quadriceps snip can be performed in combination with a more distal lateral release, provided that the superior lateral geniculate artery is preserved.

 

 

(a)

(b)

 

 

 

Figure 11.7 (a) Quadriceps snip; (b) quadriceps turndown.

 

Quadriceps turndown

This consists of an incision passing distally and laterally from the proximal extent of the standard medial parapatellar approach (Figure 11.7b). The superior lateral geniculate artery should be preserved. The inverted V thus formed can be closed as a Y, thereby advancing the quadriceps tendon and patella distally.

 

Tibial tubercle osteotomy

This requires an osteotomy of approximately 6 cm of the tibial tuberosity, hinging on the lateral soft tissues in order to maintain vascularity (Figure 11.8). The tuberosity can be proximalised in cases of patella baja. The osteotomy may be performed with a saw or sharp osteotome from the medial side and should be wide enough to include the patellar tendon insertion, tapering distally along with the anatomy of the tibial tubercle. It needs to be fixed with screws or wires at the end of the procedure.

 

Procedure

The ultimate goals of revision knee replacement are pain relief, functional stability and eradication of infection, if present. In order to achieve these goals, the important factors

 

Figure 11.8 Tibial tubercle osteotomy (leaving the lateral soft tissues undisturbed).

 

are preservation of bone stock, reconstruction of defects, adequate fixation of implants, ligamentous balancing and restoration of the joint line.

 

Implant removal

Safe and careful implant and cement removal should involve preservation of as much bone stock as possible. It is necessary to use fine, sharp osteotomes (e.g. Lambotte osteotomes), and it may be helpful to have cement-splitting osteotomes, a thin saw blade, Gigli saw and burr available. If a modular polyethylene insert is present it can be removed prior to the cemented components. It is usually preferable to remove the femoral component first as this facilitates easier extraction of the tibial component. With adequate retraction, the bone-cement interface should be carefully disrupted with osteotomes of appropriate width. If the implant is well fixed it may be safer to disrupt the implant-cement interface and remove the cement separately. Only when fully loosened should the implant be removed with the appropriate extraction device using a longitudinal distraction force. The tibial component can be removed in a similar manner. The tibial component should never be ‘levered’ out of bone. It is usually necessary to remove the cement from around the tibial keel and stem with cement-splitting osteotomes or gouges. If a polyethylene patella button has been used it should only be removed if significantly worn, in cases of infection, or if there is a patellofemoral problem. Metal-backed patella components can be very difficult to remove and are often best left if possible.

Reconstruction

Following successful removal of implants and cement, any fibrous membrane on the distal femur and proximal tibia is carefully removed with a small, sharp curette and bone nibblers. Even in cases where infection is not suspected, multiple samples should be sent

 

 

Figure 11.9 Reconstructive options for bone loss in revision knee surgery.

 

for microbiology using clean instruments. Ideally, the remaining bone surfaces should consist of trabecular bone to allow optimum cementation. Any small, contained, cavitatory defects can be filled with morsellised bone graft or cement but larger, uncontained, segmental defects need to be reconstructed with augments, wedges (Figure 11.9) or rarely, endoprosthetic replacement.

The tibia should be prepared first with the knee flexed, the proximal tibia exposed and subluxated anteriorly and the patella everted if possible. The canal is opened with a drill. Sequential reamers are used to the desired stem length until there is good endosteal engagement of the reamer. Using stable intramedullary referencing, which may be in the form of an intramedullary rod and sleeve, the proximal tibia is then resected to the correct level. If a tibial cutting jig is used with an inbuilt anteroposterior slope, the rotation should be referenced from the medial third of the tibial tubercle. Tibial resection should usually be conservative, but depends on the previous resection level and bone stock. If a flat, level cut cannot be achieved a wedge or augment may be used.

Attention is then turned to the femur. The canal is prepared as earlier. Distal femoral cuts are made with intramedullary referencing and a cutting jig. Distal femoral augments are used to prevent elevation of the joint line; if the condyles are resected at different levels, the additional resection must be compensated for by using a larger augment on that side. The femoral anterior, posterior and chamfer cuts are then performed with the cutting block positioned in the correct rotational orientation. This can be estimated from the transepicondylar axis. This is essential to ensuring a symmetrical flexion gap. Again, augments can be used to make up differences in resection levels of the posterior condyles. Thought must constantlybegivento achieving equalflexionandextensiongaps(see‘Primary total knee replacement’, p. 279) and to restoration of the joint line. If the knee is looser in flexion than extension, the femoral component is upsized with posterior augmentation. If

 

the knee is loose in extension compared with flexion, the distal femur is augmented. This is preferred to using a thicker insert and elevating the joint line. The level of the joint line should be approximately one finger breadth below the inferior pole of the patella or at the level of the meniscal scar. Femoral and tibial coverage and rotation are optimised; most systems allow the introduction of an offset between the stem and the component to facilitate this. Final trials should be performed with all trial stems, augments and wedges in place and the thickness of insert can be determined and final ligamentous releases performed. If the patella is to be revised, care must be taken with further resection. The femoral and tibial components are assembled – it can be useful to compare the completed implants to the trials to ensure no errors have been made prior to implantation. Usually, the tibia is implanted first and a separate batch of cement is used for the femur.

Two-stage revision for infection

Revising a TKR for infection is even more challenging. Two-stage revision is the current gold standard but single-stage revision is being used increasingly in cases where the soft tissues and microbiological profile allow it. Single-stage revision has better functional outcomes and exposes the patient to less surgical and anaesthetic risk. Single-stage revision should be approached in the same way as two-stage revision, with a full re-drape and change of gowns and instruments after the debridement.

At the first stage, all implants and cement are removed. Aggressive debridement is performed with excision of all infected-looking tissue, and multiple fluid and tissue samples are sent to microbiology and histopathology. The knee is thoroughly washed out. It may be helpful to perform preliminary bone cuts at this stage. A pre-moulded antibiotic-impregnated cement spacer is inserted and lightly cemented in place to avoid displacement. Appropriate antibiotics are continued and inflammatory markers checked on a regular basis. The knee should be mobilised to preserve range of motion if possible. When confident that infection has been eradicated, the second-stage revision can be performed with implantation of the definitive prosthesis. Occasionally, if not settling, the first stage may need to be repeated.

 

Closure

Routine cases can be closed in a similar fashion to primary knee replacements. Occasionally, especially after repeated revision cases or following infection, closure can be difficult and it may even be necessary to consider gastrocnemius muscle flap coverage and skin grafting, where the assistance of a plastic surgeon may be required.

 

Postoperative care and instructions

If a standard approach has been used, in aseptic cases, the postoperative regimen is similar to that following primary knee replacement. The results of microbiology samples must be obtained.

If a quadriceps turndown or tibial tubercle osteotomy has been performed, flexion should be limited for approximately 6 weeks to allow the tendon or osteotomy to heal and active quadriceps extension should be avoided.

 

Recommended references

Morgan-Jones R, Oussedik SIS, Graichen H, Haddad FS. Zonal fixation in revision knee arthroplasty. Bone Joint J. 2015;97–B:147–149.

Saleh KJ, Rand JA, Ries MD et al. Revision total knee arthroplasty. J Bone Joint Surg Am. 2003;85(Suppl 1). Younger AS, Duncan CP, Masri BA. Surgical exposures in revision total knee arthroplasty. J Am Acad

Orthop Surg. 1998;6:55–64.

Patellofemoral replacement‌

Preoperative planning

Indications

Patellofemoral replacement is indicated in the treatment of pain from isolated patellofemoral osteoarthritis when non-operative or more conservative operative management has failed.

The lateral facet of the patella and trochlea are most commonly involved, and there is commonly some degree of dysplasia, malalignment or laxity present as a predisposing factor.

Contraindications

General contraindications to knee replacement (see ‘Primary total knee replacement’, p. 279) include

 

• Tibiofemoral osteoarthritis

• Inflammatory arthritis

 

Consent and risks

• See ‘Primary total knee replacement’ (p. 279).

• Significantly higher rate of revision at 10 years compared to primary TKR (the National Joint Registry for England and Wales reports a 10-year revision rate of 18.7%).

• Specific problems with the patellofemoral articulation include patella fracture, lateral

subluxation, impingement, anterior knee pain.

 

 

Operative planning

Recent weightbearing anteroposterior, lateral and skyline radiographs must be available, and Schuss or Rosenberg views may be helpful. The Rosenberg view is a posteroanterior weightbearing view with the knee in 45° of flexion (the Schuss view is the same with the knee at 30°). Both views are more sensitive in detecting subtle tibiofemoral osteoarthritis, particularly on the lateral side. It is essential that the symptoms and signs should be consistent with patellofemoral osteoarthritis. Some surgeons consider it necessary to perform magnetic resonance imaging (MRI) or arthroscopy to assess the rest of the joint surfaces, although in most cases the decision can be made from the history, examination and plain radiographs. Occasionally, however, the final decision is made at the time of operation.

 

Choice of implants

Current designs of patellofemoral replacement are ‘onlay’ designs, and their geometry is based on existing designs of TKR. The most popular design in use in the United Kingdom, the Avon (Stryker, Newbury, United Kingdom) is based on the geometry of the Kinemax TKR and has a symmetrical trochlear groove. Others are based on more modern TKR designs and are sided. Aside from that, systems vary in their workflows, methods of femoral preparation and instrumentation. Ultimately, the decision on which implant to use rests with the surgeon and the unit.

 

Anaesthesia and positioning

Anaesthesia, positioning, preparation and draping are similar to that for primary TKR.

 

Surgical technique

Patellofemoral replacement is usually performed using a medial parapatellar approach.

 

Landmarks and incision

The position of the patella, patellar tendon and tibial tubercle should all be noted. An anterior midline longitudinal incision is made with the knee in flexion. It is not usually necessary to extend the excision as far distally as in TKR, but it needs to be long enough to allow eversion of the patella and adequate exposure of the distal femur.

 

Superficial dissection

The medial and lateral skin, subcutaneous fat and deep fascia should be reflected in a thick flap to allow exposure of the quadriceps tendon, medial patellar retinaculum and patellar tendon and to allow mobilisation of the patella.

Deep dissection

 

Structures at risk

• The anterior horns of the medial and lateral menisci should be carefully preserved, unlike with TKR where they are sacrificed. The incision at the level of the joint line must be done with great care not to extend into meniscal tissue.

• The medial femoral condyle can be damaged during the medial parapatellar approach.

• If the patellar tendon is contracted, there may be a risk of patellar tendon avulsion from the tibial tubercle during eversion of the patella. This can be prevented by extending the deep dissection proximally, dividing any lateral plicae and performing a lateral parapatellar release to allow eversion of the patella.

 

 

The medial parapatellar incision is extended from the quadriceps tendon proximally, through the medial parapatellar retinaculum and along the medial border of the patellar tendon distally. There should be an adequate cuff to ensure a good soft tissue repair. The

 

retropatellar fat pad can be incised or partially excised to facilitate eversion of the patella, and it may be necessary to perform a lateral parapatellar release. Osteophytes may be debrided at this stage.

 

Procedure

The aims of the operation are pain relief, good patella tracking and patellofemoral stability. This is achieved by accurate bone resection, correct alignment of implants and parapatellar soft tissue balancing.

Patella

To resurface the patella, the knee should be extended with the patella fully everted and held with a clamp. Peripheral osteophytes can be removed with a bone nibbler to demarcate the articular surface. It is often difficult to accurately assess the amount of patella to be resected, as the articular cartilage wear is usually not uniform. If the median ridge is of normal height, the thickness of the patella should be measured, and the amount of bone and cartilage removed should correspond to the thickness of the implant. If there is severe damage, a measured resection technique is unreliable. In this situation, the insertions of the quadriceps tendon and the lateral border of the patellar tendon can be exposed carefully with a diathermy and used as reliable landmarks. Resection 2 mm above this plane results in two-thirds of the original patella thickness being left behind. Resecting too little bone runs the risk of overstuffing the knee and if a sclerotic surface is left behind, fixation is compromised, whereas resecting too much increases the risk of fracture. A clamp is then applied and used as a cutting guide. The amount of resection should be carefully inspected and adjusted if necessary. There will usually be more bone resected from the medial than lateral facet, and the remaining cut surface may be sclerotic. This can be roughened with a saw or burr and small drill holes made to improve cement fixation; a partial lateral facetectomy may be performed if there is significant overhang. The patella is then subluxated or everted and the knee flexed to allow the femur to be prepared. Care must be taken to avoid fracture of the patella during flexion if the remaining patella is thin.

Femur

To expose the distal femur, two Hohmann retractors are placed medially and laterally. The anterior surface of the distal femur is exposed by excising the overlying synovium with coagulating diathermy. The femoral component should sit flush with the anterior distal femur without notching, with the correct degree of rotation to ensure restoration of the lateral ridge and good tracking. The implant should not be situated too far distal within the notch, as this can cause impingement and catching of the patella in full flexion. Intra- or extramedullary referencing may be used to determine the position of the implant in the coronal plane; rotation may be based on femoral landmarks (such as the transepicondylar axis or Whiteside’s line) or the long axis of the tibia, depending on the system used.

Trials are then performed and tracking assessed. Final adjustments and preparations can then be made including lateral parapatellar release if necessary. Components are cemented in place.

 

Closure

The knee should be closed in flexion in a similar manner to a primary TKR. Again, the use or not of drains is a surgical decision, but drain use is more frequently indicated if a lateral release has been performed.

 

Postoperative instructions

These are the same as for primary TKR.

 

Recommended references

Odgaard A, Madsen F, Kristensen PW et al. The Mark Coventry Award: Patellofemoral Arthroplasty Results in Better Range of Movement and Early Patient-reported Outcomes Than TKA. Clin Orthop Relat Res. 2018;476(1):87–100.

Metcalfe AJ, Ahern N, Hassaballa MA et al. The Avon patellofemoral arthroplasty. Bone Joint J. 2018;

100–B:1162–1167.

 

Unicompartmental knee replacement‌

Preoperative planning

Indications

Unicompartmental knee replacement (UKR) is indicated in the treatment of painful end-stage osteoarthritis when non-operative management has failed. The following criteria must be met:

• Bone-on-bone osteoarthritis or osteonecrosis in a single compartment

• Presence of full-thickness cartilage in the remaining compartment

• Varus or valgus deformity must be correctable to normal

• Fixed flexion deformity less than 10°

• Functionally intact knee ligaments

   

  Contraindications

• General contraindications to knee replacement (see ‘Primary total knee replacement’, p. 279).

• Inflammatory arthritis.

• Relative contraindications are controversial. Some authors (most notably the

  Oxford group) recommend offering UKR to all patients in whom the pathoanatomy is suitable, with no restrictions based on patient factors. Others have suggested contraindications including obesity, youth, high activity levels, the presence of chondrocalcinosis, the presence of generalised knee pain or the presence of a lateral osteophyte (in medial UKR). A number of clinical studies from various centres have been performed which suggest that these factors are unimportant in determining outcome following UKR.

• Patellofemoral osteoarthritis is particularly controversial, and some surgeons suggest strongly that UKR should not be performed in patients with PFJ disease. Again, there are several studies demonstrating no effect of clinical or radiological evidence of patellofemoral osteoarthritis (aside from severe lateral wear) on outcome.

 

 

Consent and risks

• Most risks and complications of primary TKR can occur in UKR.

• Medial knee pain may occur and usually resolves with time. Persistent anteromedial knee pain may be associated with changes in loading of the medial bone or damage to the MCL intraoperatively.

• The rate of revision of UKR is two to three times that of TKR. This is partly due to additional

  mechanisms of failure (including progression of osteoarthritis) and partly due to a lower threshold for revision of UKR. High-volume centres report revision rates similar to those of TKR.

• In common with TKR, most patients undergoing UKR never undergo revision surgery.

  Outcomes of revision of UKR to TKR are closer to the outcomes of a revision TKR than to those of a primary TKR. For these reasons UKR should be performed in those with end-stage disease and should not be considered as a temporising measure for TKR.

• Dislocation of the insert can occur with mobile bearings, especially if ACL laxity is present.

• Revision of UKR usually results in a ‘primary’ knee replacement but bone loss around the tibial baseplate may necessitate stems and wedges.

 

 

Operative planning

Recent weightbearing anteroposterior, lateral and skyline radiographs must be available, and Rosenberg views can be helpful in identifying subtle disease in the preserved compartment. Some surgeons advocate the use of valgus and varus stress views to determine suitability for UKR. Anteromedial arthritis is the most common indication; the presence of an anterior wear scar on the lateral radiograph implies the presence of an intact ACL (Figure 11.10). Some surgeons advocate the use of MRI or arthroscopy to determine eligibility for UKR, but in most cases the decision can be made by the history, examination and plain radiographs. However, the final decision is made at the time of operation.

 

 

(b)

(a)

 

 

 

Figure 11.10 Anteromedial arthritis. Note the anterior wear scar on the lateral radiograph.

 

Choice of implant

All implants in UKR have the same aim, which is to restore the affected joint surface and recreate native kinematics by reconstructing the joint with the minimum of constraint. To do this there are two main design rationales. Most designs of UKR have a fixed bearing, flat tibial component, which may be all-polyethylene or metal-backed, and a polyradial femoral component. The most commonly used UKR, the Oxford Knee (Zimmer Biomet, Bridgend, United Kingdom) uses a different rationale, with a spherical femoral component and a mobile polyethylene bearing, which has a highly conforming articulation with the femur on the superior surface and a flat-on-flat articulation with the metal tibial base plate on the underside. Proposed advantages of the mobile bearing include improved wear properties and kinematics, but this comes with a risk of dislocation of the bearing, which occurs in just under 1% of cases in large series. Most designs of UKR are cemented but cementless fixation is becoming increasingly popular and has excellent published results.

 

Anaesthesia and positioning

This is essentially similar to that described for TKR. However, as UKR is usually performed via a less invasive approach, the use of regional anaesthesia can be avoided, if desired, by the administration of local anaesthetic into the wound and deep tissues. Some surgeons prefer to use a leg holder with the knee flexed, the hip abducted and the leg over the side of the operating table or the foot of the table removed. This allows the knee to be stressed and can improve exposure.

 

Surgical technique

Medial unicompartmental knee replacement

Landmarks and incision

With the knee flexed, a longitudinal incision is made along the medial border of the patellar tendon from patella to tibial tubercle and can be extended proximally or distally as required.

 

Dissection

The incision is deepened along the same line and the medial border of the patella and tendon identified. It is continued along the medial border of the patellar tendon and proximally up to the medial parapatellar retinaculum. The medial capsule is dissected subperiosteally off the proximal tibia to gain exposure to the medial compartment. The dissection should not extend beyond the anteromedial corner and should not involve any release of the MCL. The medial portion of the fat pad can be excised. The anterior two-thirds of the medial meniscus can be excised at this point, with the posterior horn removed later following bone cuts. This should give adequate exposure of the medial compartment, and it should be possible to inspect the ACL, patellofemoral joint and lateral compartment.

This operation can usually be performed through a relatively minimally invasive approach, with the skin incision being used as a ‘mobile window’ to gain access to the femur or tibia with varying degrees of knee flexion. However, if exposure is difficult, the skin incision and deep dissection should be extended to allow the patella to be subluxated laterally, although it should not usually be necessary to involve the quadriceps tendon or vastus medialis.

 

Procedure

 

Structures at risk

• The MCL must be protected throughout.

• The ACL is at risk during the sagittal tibial cut with the reciprocating saw and should be retracted.

• The patellar tendon can be damaged during reaming of the femoral condyle.

 

 

Osteophytes on the medial tibial plateau and femoral condyle are excised. The exact nature and sequence of bone preparation are dependent on the implant used and manufacturer’s recommendations but the aims are restoration (or slight undercorrection) of pre-disease alignment, equal flexion/extension gaps, optimum range-of-motion and good fixation of implants.

The tibial cut is usually made first, with extramedullary referencing and a tibial cutting guide. The vertical cut is performed, with a reciprocating saw, just medial to the ACL insertion. The horizontal cut is made with an oscillating saw, perpendicular to the long axis of the tibia. The wedge of tibia can then be removed with a Kocher forceps.

The femoral preparation uses femoral intramedullary alignment and the tibial cut as a combined reference (Figure 11.11). The flexion gap is usually set by making the posterior condylar cut first. The mechanism of setting the extension gap varies by implant and can involve saw cuts or reaming.

Final trials and preparations can then be made and the definitive implants inserted. If there is impingement of the bearing anteriorly on the femoral condyle in full extension, trimming of the condyle can be performed to allow clearance.

Closure

Closure is in layers, with continuous absorbable sutures and a continuous subcuticular suture or staples to the skin. It is not usually necessary to use a drain.

 

Lateral unicompartmental knee replacement

This is much less commonly performed than medial UKR. It can be performed via either a midline approach with the patella everted, or a direct lateral parapatellar approach. The procedure itself is analogous to that of medial unicompartmental surgery, although some surgeons use a window in the patellar tendon to perform the vertical saw cut. Due to the increased excursion of the lateral compartment during knee movement, and the relative laxity of the lateral compartment in flexion the rate of bearing dislocation is significantly higher in mobile bearing UKR, and fixed bearings are often used.

 

Postoperative care and instructions

Patients should be encouraged to mobilise the knee and bear weight as quickly as possible. Recovery is significantly faster than after TKR and patients leave hospital on average a day earlier. The use of day-case UKR is increasing around the world.

 

Figure 11.11 Component alignment in unicompartmental knee replacement.

 

Recommended references

Argenson JN, Blanc G, Aubianiac JM, Paratte S. Modern unicompartmental knee arthroplasty with cement: A concise follow-up, at a mean of twenty years, of a previous report. J Bone Joint Surg Am. 2013:95:905–909.

Liddle AD, Judge A, Pandit H, Murray DW. Adverse outcomes after total and unicompartmental knee replacement: A study of 101,330 matched patients from the National Joint Registry for England and Wales. Lancet. 2014;384:1437–1445.

Pandit H, Jenkins C, Gill HS et al. Unnecessary contraindications for mobile-bearing unicompartmental knee replacement. J Bone Joint Surg Br. 2011;93:622–628.

 

Distal femoral osteotomy‌

Preoperative planning

Distal femoral osteotomy is used to correct valgus deformity of the knee and consists of a varus osteotomy which may either be a medial closing wedge or a lateral opening wedge. The authors’ preference is the lateral opening wedge varus osteotomy and this is described next.

 

Indications

Distal femoral osteotomy is indicated in the treatment of pain and deformity caused by valgus osteoarthritis in relatively young patients when non-operative management has failed.

It can also be used to correct malunion following supracondylar fractures of the femur.

 

Contraindications

• Distal lower limb ischaemia

• Significant medial or patellofemoral osteoarthritis

• Flexion limited to less than 90°

• Fixed flexion deformity greater than 15°

• Inflammatory arthritis

• Osteoporosis

• Inability to comply with the rehabilitation protocol

 

 

Consent and risks

• Delayed/non-union

• Inadequate/loss of correction

• Failure of fixation

• Stiffness

• Iliotibial band irritation

• Progression of arthritis may require revision to TKR. Although this can be relatively straightforward, it may be advisable to remove the metalwork at a separate operation prior to performing TKR

• There may be difficulty in achieving the desired valgus intramedullary alignment of the

  distal femoral bone cut following distal femoral osteotomy

• Infection

• Bleeding

• Venous thromboembolism

• Neurovascular injury

 

 

Operative planning

Recent weightbearing anteroposterior, lateral and skyline radiographs must be available. Long-leg alignment films must be performed and stress views may be helpful. It is essential that the symptoms and signs should correlate with radiographic findings. It is sometimes necessary to perform MRI or arthroscopy to assess the integrity of the ligaments and the state of the joint surfaces. Although instability has historically been thought of as a contraindication to osteotomy, it may be performed as a precursor to or in association with ligament reconstruction in malaligned ligament deficient knees. Accuracy of correction is of paramount importance and the osteotomy must be planned using templating software; there may be a role for patient-specific instrumentation. The rule of thumb is that each millimetre of opening in the coronal plane corresponds to 1° of correction.

 

Anaesthesia and positioning

General anaesthesia is used. Regional anaesthesia can be avoided, if desired, by the administration of local anaesthetic into the wound and deep tissues. The patient is positioned supine on the operating table. Provided that there are no contraindications, a tourniquet is applied as proximally as possible. An image intensifier is needed throughout the operation.

Prior to preparation and draping, in order to reference the mechanical axis of the limb, the centre of the femoral head can be screened with the image intensifier and a radio-opaque electrocardiogram (ECG) sticker is placed on the skin directly overlying the centre of the femoral head. If a tri-cortical wedge of iliac crest bone graft is to be used, the iliac crest must be prepared, draped and exposed to allow bone graft harvesting.

Surgical technique

The aim of a varus osteotomy is to correct valgus malalignment, shift the mechanical axis to the medial compartment and offload the diseased lateral compartment (Figure 11.12). The goal is to overcorrect to a tibiofemoral angle of 0°. Traditionally, a distal femoral osteotomy is preferred for varus osteotomy as this more reliably achieves a horizontal joint line.

 

Figure 11.12 Lateral opening wedge distal femoral osteotomy in a valgus knee.

 

Landmarks and incision

An opening wedge osteotomy is performed via a lateral approach to the distal femur. A longitudinal incision is made over the lateral aspect of the thigh in the supracondylar region. Correct placement of the incision site is ensured by screening with the image intensifier.

 

Dissection

The incision is deepened along the same line until the fascia lata is exposed. The fascia lata is split and the vastus lateralis can either be divided or incised and lifted off the femur at its posterior border. Any blood vessels encountered should be coagulated and subperiosteal dissection is continued anteriorly and posteriorly around the femur. Insertion of the appropriate retractors anteriorly and posteriorly gives good exposure of the distal femur.

 

Procedure

 

Structures at risk

• The popliteal artery must be protected by the subperiosteal retractor throughout the procedure. The risk of major arterial injury may be reduced by performing the osteotomy with the knee in flexion as this moves the artery away from the posterior femur.

• The medial distal femoral cortex should be left intact. If breached, a staple can be inserted

to maintain stability.

 

 

The ECG sticker placedover thefemoralhead ispalpatedthroughthedrapesandanalignment rod can be placed to lie between this point and the centre of the articular surface of the ankle joint. When the osteotomy is opened by the correct amount, the mechanical axis is shifted to the desired point, i.e. the centre of the medial tibial plateau. This improves the efficiency of the use of fluoroscopy, minimises X-ray exposure and helps to minimise operation time.

A first guide wire is inserted with the power driver from the lateral cortex in a medial and slightly caudal direction. The wire should emerge in the metaphyseal region of the distal femur at the junction of the medial femoral condyle and the supracondylar ridge. A second wire is then introduced to lie exactly superimposed on the first true anteroposterior fluoroscopic image, indicating that the wires are exactly parallel to the joint surface. This second wire can be introduced through a parallel guide.

The osteotomy can then be performed with an oscillating saw, using either the guide wires or cutting jig to help control the saw. The saw is placed on the proximal side of the wires and advanced approximately two-thirds of the distance across the femur under fluoroscopic control. Care must be taken not to penetrate the medial cortex. The osteotome is then used to complete the osteotomy through the anterior and posterior cortices but should stop approximately 1 cm short of the medial cortex. The blade of the osteotome can be marked at a level where penetration of the far cortex will not occur, and this marking can be observed carefully as the osteotome advances.

The osteotomy is then opened with distraction osteotomes using a screwdriver under fluoroscopic control. When the osteotomy is opened, metal wedges can be gently inserted into the osteotomy to the desired level. The amount of correction can be checked with the image intensifier using the alignment rod as previously described. A locking plate with interposition wedge of the desired size is then inserted into the osteotomy in the correct position and screws inserted and checked with fluoroscopy. The opening wedge can be filled with bone graft or calcium triphosphate wedges.

 

Closure

Closure is performed in layers with continuous absorbable sutures and a continuous subcuticular suture or staples to the skin. It is not usually necessary to use a drain.

 

Postoperative care and instructions

Regular neurovascular observations should be performed and the patient carefully monitored for signs of compartment syndrome. Adequate analgesia is administered. Mechanical and chemical thromboprophylaxis is recommended. Two further doses of prophylactic antibiotics are administered at 8 hours and 16 hours postoperatively. The wound should be inspected and radiographs performed prior to discharge should be checked. If the fixation is stable, range-of-motion exercises are encouraged from the first postoperative day. Patients should remain non-weightbearing in a hinged knee brace for 2 weeks. Repeat radiographs are taken and clips removed at this stage. Touch weightbearing only is commenced in a hinged knee brace for a further 4 weeks. If radiographs are satisfactory at 6 weeks, partial weightbearing can be commenced and if the osteotomy has united at 12 weeks the patient can build up to full weightbearing.

 

Recommended references

Brouwer RW, Raaij van TM, Bierma-Zeinstra SM et al. Osteotomy for treating knee osteoarthritis.

Cochrane Database Syst Rev. 2007;(18):CD004019.

Cameron JI, McCauley JC, Kermanshahi AY, Bugbee WD. Lateral opening-wedge distal femoral osteotomy: Pain relief, functional improvement, and survivorship at 5 years. Clin Orthop Relat Res. 2015;473:2009–2015.

Wylie JD, Scheiderer B, Obopilwe E et al. The effect of lateral opening wedge distal femoral varus osteotomy on tibiofemoral contact mechanics through knee flexion. Am J Sports Med. 2018;46:3237–3244.

 

Proximal tibial osteotomy‌

Preoperative planning

Proximal tibial osteotomy is used to correct varus deformity of the knee and consists of a valgus osteotomy which may either be a lateral closing wedge or a medial opening wedge. The authors’ preference is the medial opening wedge valgus osteotomy; this is described later.

 

Indications

Proximal tibial osteotomy is indicated in the treatment of pain and deformity caused by varus osteoarthritis in relatively young patients when non-operative management has failed.

Contraindications

• Distal lower limb ischaemia

• Significant lateral or patellofemoral osteoarthritis

• Significant bone loss from the medial tibial plateau

• Flexion limited to less than 90°

• Fixed flexion deformity greater than 15°

• Inflammatory arthritis

• Osteoporosis

• Inability to comply with rehabilitation protocol

 

 

Consent and risks

Progression of arthritis may require revision to TKR. Although this can be relatively straightforward, it may be advisable to remove the metalwork prior to performing TKR. There may be problems caused by patella baja. Other specific complications include

• Infection

• Bleeding

• Venous thromboembolism

• Common peroneal nerve injury (usually associated with fibular osteotomy in closing wedge proximal tibial osteotomy)

• Major arterial injury

• Compartment syndrome

• Lateral tibial plateau fracture

• Delayed/non-union

• Inadequate/loss of correction

• Overcorrection

• Failure of fixation

• Stiffness

 

 

Operative planning

This is the same as for distal femoral osteotomy.

 

Anaesthesia and positioning

This is the same as for distal femoral osteotomy.

 

Surgical technique

The aim of a valgus osteotomy is to correct malalignment, shift the mechanical axis to the lateral compartment and offload the diseased medial compartment (Figure 11.13). The goal is to correct to a tibiofemoral angle of 5°–9°. The advantages of an opening wedge valgus osteotomy include the fact that there is no need for a fibular osteotomy, there is more control over the correction and it may correct instability in anterior or posterior cruciate ligament deficiency by adjustment of the tibial slope. It has the disadvantage of creating a degree of patella baja.

 

Landmarks and incision

A medial opening wedge valgus osteotomy is performed via an anteromedial approach to the proximal tibia. A longitudinal or oblique incision is made over the anteromedial aspect of the proximal lower leg in the region of the insertion of the pes anserinus and 3 cm medial to the lower border of the tibial tubercle.

 

 

Figure 11.13 Medial opening wedge proximal tibial osteotomy in a varus knee.

 

Dissection

The incision is deepened until the fascia overlying the pes is exposed. The fascia is incised and the pes anserinus is reflected posteriorly, with the superficial medial collateral ligament. Anybloodvesselsencounteredshouldbecoagulatedandsubperiostealdissection is continued anteriorly and posteriorly around the tibia. Insertion of the appropriate retractors anteriorly and posteriorly gives good exposure of the proximal tibia and protects the patellar tendon anteriorly with the popliteal artery and tibial nerve posteriorly.

 

Procedure

 

Structures at risk

• Popliteal artery: The risk of major arterial injury may be reduced by performing the osteotomy with the knee in flexion.

• Patellar tendon.

• Superficial MCL.

• The lateral tibial cortex should be left intact. If breached, a staple should be inserted laterally to maintain stability.

• The lateral tibial plateau can be fractured if the anterior cortex has not been fully

osteotomised prior to distraction of the osteotomy. If this occurs, the osteotomy should be advanced and the fracture stabilised with one or more interfragmentary screws from the lateral side.

 

 

The ECG sticker placed over the femoral head is palpated through the drapes and an alignment rod can be placed to lie between this point and the centre of the articular surface of the ankle joint. When the osteotomy is opened by the correct amount, the mechanical axis is shifted to the desired point, i.e. at the junction of the medial two-thirds and lateral third of the articular surface of the tibia. Using an alignment rod to show the mechanical axis improves the efficiency of the use of fluoroscopy, minimises X-ray exposure and helps to minimise operation time. It is imperative that a ‘true’ anteroposterior radiograph of the knee be obtained in order to gauge the anteroposterior slope of the tibia.

A first guide wire is inserted, with the power driver, from the medial cortex in a lateral and slightly cephalad direction. The wire should emerge in the metaphyseal region of the proximal tibia at the level of the tip of the fibula head. A second wire is then introduced to lie exactly superimposed on the first on a true anteroposterior fluoroscopic image, indicating that the wires are exactly parallel to the joint surface. This second wire can be introduced through a parallel guide.

The osteotomy can then be performed with an oscillating saw, using either the guide wires or a cutting jig to help control the saw. The saw is placed on the distal side of the wires and advanced approximately two-thirds of the distance across the tibia under fluoroscopic control. Care must be taken not to penetrate the lateral cortex. The osteotome is then used to complete the osteotomy through the anterior and posterior cortices but should stop approximately 1 cm short of the lateral cortex. The blade of the osteotome can be marked at a level where penetration of the far cortex will not occur, and this marking can be observed carefully as the osteotome advances.

The osteotomy may pass above the insertion of the patellar tendon, but if it crosses the anterior tibial cortex at the level of the tibial tuberosity it may be necessary to make a step cut beneath the tuberosity from the transverse osteotomy proximally to ensure that the tuberosity and patellar tendon insertion remain intact (Figure 11.14).

The osteotomy is then opened with distraction osteotomes using a screwdriver under fluoroscopic control. When the osteotomy is opened, metal wedges can be gently inserted into the osteotomy to the desired level. The amount of correction can be checked with the image intensifier using the alignment rod as previously described. A locking plate with interposition wedge of the desired size is then inserted into the osteotomy in the correct position and screws inserted and checked with fluoroscopy. The opening wedge can be filled with bone graft or calcium triphosphate wedges. Autograft is still advocated for larger corrections or revision procedures.

 

Closure

Closure is undertaken in layers with continuous absorbable sutures and a continuous subcuticular suture or staples to the skin. It is not usually necessary to use a drain.

 

Postoperative care and instructions

These are the same as for distal femoral osteotomy.

 

 

Step cut to avoid tubercle

 

 

 

Figure 11.14 A step cut to avoid tibial tuberosity.

 

Recommended references

Brinkman JM, Lobenhoffer P, Agneskirchner JD et al. Osteotomies around the knee: Patient selection, stability of fixation and bone healing in high tibial osteotomies. J Bone Joint Surg Br. 2008;90:1548–1557. Coventry MB, Ilstrup DM, Wallrichs SL. Proximal tibial osteotomy. A critical long-term study of eighty-

seven cases. J Bone Joint Surg Am. 1993;75:196–201.

Niinimäki TT, Eskelinen A, Mann BS, Junnila M, Ohtonen P, Leppilahti J. Survivorship of high tibial osteotomy in the treatment of osteoarthritis of the knee: Finnish registry-based study of 3195 knees. J Bone Joint Surg Br. 2012;94:1517–1521.

 

Knee arthrodesis‌

Preoperative planning

Indications

The most common indication for knee arthrodesis is as a salvage procedure in failed revision knee replacement either where infection has been resistant to eradication or when there is a non-functioning extensor mechanism.

In the past and currently still in some parts of the world, knee arthrodesis is more commonly performed in patients with postinfective arthritis, tuberculosis, poliomyelitis and severe trauma.

Contraindications

• Critical arterial ischaemia

• Extensive bone loss (relative)

• Ipsilateral hip arthrodesis (relative)

 

 

Consent and risks

• Infection (or failure to eradicate existing infection)

• Bleeding

• Venous thromboembolism

• Wound problems

• Neurovascular injury

• Fractures

• Delayed or non-union

• Pain

• Immobility

• Risk of subsequent amputation

• Increased risk of osteoarthritis in other joints of the lower limbs

 

 

Operative planning

The indication for arthrodesis, severity of bone loss and adequacy of soft tissue coverage all need to be taken into account prior to deciding on whether the arthrodesis should be a single or staged procedure or an intramedullary or extramedullary fixation and whether it is necessary to enlist the help of a plastic surgeon. The patient should be thoroughly counselled before the operation.

Anaesthesia and positioning

See ‘Revision total knee replacement’ (p. 293).

 

Surgical technique

This depends on indication, type of fixation and need for bone grafting.

 

Landmarks and incision

A longitudinal, anterior midline incision is used, usually through a previous TKR scar.

 

Dissection

The quadriceps tendon and patellar tendon are split and a patellectomy performed. A synovectomy can be carried out and ligaments released or divided to gain exposure.

Procedure

The goals of arthrodesis are pain relief, eradication of infection and sound bony fusion in the correct alignment. In order to achieve these goals the important factors are good

 

apposition of healthy bone surfaces, preservation of bone stock and stable fixation with compression.

Implants are removed as described in the section ‘Revision total knee replacement’ (p. 293). It is important to preserve as much bone as possible. Bone surfaces must be viable. In the presence of infection, it is usually desirable to perform a two-stage procedure with the first stage involving thorough debridement, insertion of an antibiotic-impregnated cement spacer and temporary fixation. The second stage involves the definitive arthrodesis. This is commonly achieved in one of two ways.

External fixation

The tibia is cut perpendicular to the long axis. The femur is cut to enable apposition at approximately 15° of flexion, 7° valgus and 10° external rotation. Bone graft may be used if desired. Compression must be achieved with the external fixator. Any form of external fixator can be used, from simple monoaxial fixators to fine wire frames.

Intramedullary fixation

Bone cuts are made as described earlier. The femoral and tibial intramedullary canals are reamed. The distal femur/proximal tibia can be reamed in a concave/convex fashion to increase contact surface area. Fixation can be achieved either with a long nail or with a two-part nail with a locking device between the femur and tibia which can also provide compression, correct alignment and restore length in cases with significant bone loss. The nail can be locked proximally and distally.

 

Closure

In some cases closure can be difficult. Occasionally, especially after repeated revision cases or following infection, closure can be such a challenge that it may even be necessary to consider gastrocnemius muscle flap coverage and skin grafting, where the assistance of a plastic surgeon may be required.

 

Postoperative care and instructions

The amount of weightbearing allowed depends on the stability of fixation, but generally touch weightbearing should be commenced immediately, gradually built up to partial weightbearing over approximately 6 weeks and to full weightbearing over the next 6 weeks.

 

Recommended references

Conway JD, Mont MA, Bezwada HP. Arthrodesis of the knee. J Bone Joint Surg Am. 2004;86:835–848.

White CJ, Palmer AJR, Rodriguez-Merchan EC. External fixation vs intramedullary nailing for knee arthrodesis after failed infected total knee arthroplasty: A systematic review and meta-analysis. J Arthroplasty. 2018;33:1288–1295.

Wiedel JD. Salvage of infected total knee fusion: The last option. Clin Orthop Relat Res 2002;(404):139–142.

 

 

Viva questions‌

1. What are the risks and complications of TKR?

2. What are the contraindications to TKR?

3. Which TKR would you choose and why?

4. Discuss the advantages and disadvantages of posterior cruciate ligament retaining and sacrificing TKR.

5. Describe how you would address an imbalance in flexion/extension gaps.

6. What releases would you perform to correct alignment in a valgus knee?

7. How do you deal with a fixed flexion deformity during TKR?

8. What measures do you take to ensure correct patella tracking in primary TKR?

9. How would you manage a patient with a painful knee replacement?

10. Describe the modes of failure of TKR.

11. What extensile approaches are available in revision knee replacement?

12. What is your rationale for choosing an implant in revision knee replacement?

13. What are the treatment options for a 50-year-old man with medial compartment osteoarthritis?

14. What criteria need to be met for a medial UKR?

15. How would you select the ideal patient for a patellofemoral replacement?

16. How would you ensure adequate realignment in proximal tibial osteotomy?

17. Why is a distal femoral osteotomy preferred to proximal tibial osteotomy in a valgus knee?

18. Discuss the pros and cons of opening versus closing wedge proximal tibial osteotomy.

19. What are the indications for knee arthrodesis?

20. Describe the general principles and fixation options in arthrodesis.