Revision total knee replacement- First stage of Implantcast MUTARS MK EPR using reinforced static antibiotic cement spacer

AP radiograph prior to the first-stage revision of an infected revision hinged knee prosthesis with gross metaphyseal femoral bone loss, impending fracture and lysis of the proximal tibia.
Pre-operative aspiration had identified staph.epidermidis species which was sensitive to a variety of antibiotics. After MDT discussion, the decision was taken to offer staged revision because this was a re-revision for recurrent PJI and because of the expected challenges in explanting the trial prosthesis.
It was also felt that the residual bone stock of the distal femoral metaphysis was inadequate and that the second stage reconstruction would involve a distal femoral endoprosthesis.

The limb is isolated after a tourniquet is applied and two-stage skin preparation is commenced using alcoholic chlorhexidine.

After applying a ‘bowel bag’, used in general sugery, to the foot, the isolated limb is covered with Ioban incisional drapes as shown. Routine props for knee revision arthroplasty are placed lateral to the thigh and to rest the foot distally.

Skin incision with the knee flexed along the previous incision.It is important to use previous incisions as much as possible to minimise the risk of skin necrosis.
As with all revision arthroplasty, extending the previous incisions longitudinally helps to identify the tissue layers for dissection.

Using McIndoe’s dissecting scissors and/or diathermy, full thickness skin flaps are elevated above the capsular layer of the knee joint.This is to minimise the risk of skin necrosis, which in multiply revised joints is often a well defined, but thin, layer of tissue, as this photograph highlights.

Prior to the arthrotomy, aspirate the joint for the first of five microbiological samples submitted to the laboratory.The fluid is placed into blood culture bottles for culture and sensitivities. The other four microbiology samples are all representative tissue samples from the joint and femoral/tibial medullary cavities later in the procedure.
The needle is often useful to define the medial border of the patella, which often broadens and flattens in previously revised cases with unresurfaced patellae, making identification of the medial parapatellar arthrotomy challenging.

The arthrotomy is performed using the previous arthrotomy, most likely medially, but always useful to check the pre-operative deformity and/or operation notes to be sure.I would always ensure that sufficient soft-tissue is left around the patella to ensure a sound closure at the end of the procedure.

Immediate appearances are of grossly infected polyethylene debris lining the joint with staining of the tissues. All of this tissue will need to be physically removed to optimise our chances of PJI eradication.

Part of the polyethylene patella button was free floating in the supra-patellar pouch and removed.

Using Lane’s soft-tissue holding forceps for traction and McIndoe’s dissecting scissors, identify the correct layer for a complete synovectomy of the supra-patellar pouch and medial and lateral gutters.Remove all the tissue contaminated by the infection down to the quadriceps tendon and muscles and bone. Any residual infectious material could lead to recurrent PJI, which in this context has a high risk of amputation.

Once the layer is identified, with appropriate counter traction using Kocher’s forceps and a nibbler, monopoly diathermy can be used to develop this plane leaving healthy bleeding tissue on the side of the Lane’s soft-tissue holding forceps.
A radical debridement is thought to be key to eradicating PJI; a controlled but complete, rather than cautious approach to the debridement is required.

Continue the anterior synovectomy around the patella, which has the residual patella button still attached.
The patella is a good place to help identify the correct plane for debridement to avoid over resecting tissue from the patella and quadriceps tendons by identifying the level of the previous patella resurfacing osteotomy.

Having cleared the tissue around the patella, the plane of dissection flows into the lateral gutter. As more and more dissection occurs the patella will become easier to evert, maximising access to the joint. This is why doing the synovectomy first is advantageous.

With the knee now flexed, the anterior synovectomy can flow distally via the medial and lateral gutters down to the tibia.
I find using the heavy nibblers to grasp the tissue being excised maintains sufficient tension to almost peel this layer off the femur and capsule, using the diathermy just to keep it flowing.

We have now cleared the synovectomy over the medial femoral condyle (MFC) to the medial proximal tibia, where there is gross (O) osteolysis, shown by the fibrinous “creamy cheese” material where the proximal tibia previously existed.
All of the tissue from the ‘effective joint space’ requires radical excision. Thankfully the bone looks a little more supportive at the tibial tubercle.

Having cleared the anterior part of the joint this single piece of tissue is excised and sent for histopathology and microbiology to confirm the presence of infection and the causative organisms and their sensitivities.

Explant the polyethylene using a screwdriver to back out the screw that locks the poly into the tibial prosthesis.

Once the screw has been removed, the polyethylene prosthesis is easily lifted out of the wound.

Explant the femoral prosthesis.We now have complete exposure of the femoral prosthesis having cleared the (M) medial and (L) lateral gutters of the thick infected synovium and removing the polyethylene.
A (H) Heygroves bone holding forceps is used to grasp the femur proximally in preparation for the inevitable transfemoral osteotomy because the bone quality is so poor.

As the X-rays indicated, the femoral prosthesis is so loose that it was simply lifted out by hand revealing gross femoral bone loss (F).

There are various classifications for categorising femoral bone loss.
I prefer the AORI classification (Engh GA, Ammeen DJ. Bone loss with revision total knee arthroplasty: defect classification and alternatives for reconstruction. Instructional Course Lectures. 1999 ;48:167-175), as is simple and logical.
This would be a F3 using Engh’s classification as there is an uncontained defect requiring endoprosthetic replacement: http://eknygos.lsmuni.lt/springer/238/116-132.pdf.

In order to resect the distal femur it is necessary to release the gastrocnemius origins off the posterior femur to skeletonise the bone prior to osteotomy.Using a bone hook placed into an osteolytic defect laterally to elevate the femur, cautery is use to develop the subperiosteal plane and divide the muscle origins off the posterior femur. At this point the popliteal artery, vein and tibial nerve are deep to the plane of dissection, so care must be taken at this stage.

As the femur is elevated out of the wound, the dissection of the posterior femur is made easier and safer as the popliteal vein, artery and tibial nerve lie deep to the femur in the popliteal fossa: at the level of the knee joint the artery lies closest to our plane and slightly more medial than the vein and lastly the most superficial and lateral tibial nerve. Far laterally lies the peroneal nerve (although at this point should be deep to the biceps femoris tendon).

The posterior dissection and gastrocnemius release off the distal femur proceeds to the level of the pre-operatively planned osteotomyThis is above the level of the previous femoral stem where the bone appears more normal.
The proximal femur is held with the Heygroves bone folding forceps and a 1.27mm oscillating saw blade used to divide the femur.
To protect the neuromuscular structures beneath, a large (L) Langenbach’s retractor is placed underneath the femur from the opposite side to the surgeon, as shown, because at this level the vessels exiting the adductor canal lie posteromedial to the femoral shaft, beneath the adductor magnus tendon. The contents of the adductor canal (Hunter’s canal/subsartorial canal)) include the femoral artery, femoral vein, nerve to the vastus medialis and the saphenous nerve. The femoral vessels become the popliteal vessels as they exit the canal and lie in the popliteal fossa of the posterior distal thigh. The saphenous nerve arises from the femoral nerve (L3,4) and once it reaches the posterior thigh becomes subcutaneous by piercing the fascia between the sartorius and gracilis tendons and runs with the great saphenous vein behind the medial border of the tibia to the foot and ankle.

The resected distal femur is removed from the wound and can be sent for histology or microbiology if required.

Posterior synovectomy is completed using cautery down to the posterior tibial plateau.Having removed the distal femur, the huge exposure to the wound permits completion of the synovectomy posteriorly.
Again, maintain tension using the nibblers, diathermy is used to resect the layer of infected synovial tissue.
At this stage the popliteal artery is the closest neurvascular structure centrally, then the popliteal vein, then the tibial nerve in the popliteal fossa. Consequently, great care is taken to do this thoroughly and radically without damaging these structures. Usually the layer is identifiable from the edges of the previous anterior synovectomy, but it can be challenging after segmental bone resection to appreciate exactly the relations of critical structures once the bone landmarks have been excised.

As before, the tissue being excised is retracted out of the wound anteriorly, to permit access to the (PT) posterior tibia.
As the posterior synovectomy travels proximally to distally, flexing the ‘knee’ helps the surgeon to get underneath the thick synovium and hopefully keeps the vessels out of danger, as they remain below the level of the tibial metaphysis, immediately posterior and slightly medial to the centre of the tibial plateau.

The posterior synovectomy is completed, and as this photograph highlights, once the synovectomy is complete and the knee is flexed, we have circumferential exposure of the proximal tibia.

Explant the proximal tibial prosthesis using osteotomes and revision arthroplasty hand instruments.I would start by identifying the prosthesis – cement interface and attempt to disrupt this using osteotomes. Because of the loss of the distal femur, the assistant apples counter resistance to blows from the (O) osteotome.
I normally start in the centre aiming for the central tibial stem. By using the prosthesis-cement interface rather than the cement – bone interface, in theory one avoids damaging the residual bone. In this case though the medial bone loss is severe already due to the polyethylene debris induced osteolysis and infection.

Once the plane between the prosthesis and cement has been established, circumferentially complete the cement disruption, here using the excellent exposure following removal of the distal femur to attack the posterior tibial interface using a broad flat (O) osteotome and (M) mallet.

Once the prosthesis – cement interface has circumferentially been disrupted (indicated here by the gap now visible beneath the tibial tray, we might attempt tibial explant using an instrument we call a ‘coal chisel’ but is effectively a large punch for impacting prostheses.

On this occasion I have been exceptionally lucky and the tibial prosthesis lifts out of the tibia with ease.

Remove the cement from the proximal tibia.To complete the radical debridement, all implanted material exposed to the PJI, including the cement must be removed.
I start by using the deficient medial proximal tibial metaphysis to remove the cement beneath the tibial tray using osteotomes and nibblers.

A Cobb elevator is used to lift the cement off the residual bone without causing further bone loss or damage.

Cement splitters are used to remove the cement that is fixed to the residual host bone. This requires controlled aggression and patience. Always give the cement space to fracture into, use single sharp blows rather than repetitive banging, and never twist well fixed cement for fear of fracturing the host bone! The hand cement instruments used mostly include the cement splitters (in three different sizes), cement gouges and chisels, the cement grasping forceps and the retrograde cement chisels (‘back-scratchers’) found on any revision arthroplasty set.

Having removed the last of the cement in the proximal tibia it’s obvious there is severe bone loss on the medial proximal tibia.This would be a T3 using Engh’s classification as there is an uncontained defect.
This will need a cone and long-stem fixation at the time of reconstruction which I record on the operation note to remind me when planning the second stage
Engh GA, Ammeen DJ. Bone loss with revision total knee arthroplasty: defect classification and alternatives for reconstruction. Instructional Course Lectures. 1999 ;48:167-175

Viewed from the front, the medial bone loss is evident, but the (TT) tibial tubercle remains intact with an intact extensor mechanism attached.

Debride the femoral canal using hand revision instruments.Although the distal femoral resection removed most of the foreign material on the femoral side of the joint, there remains some residual medullary cement which requires extraction.
Options include cement reamers, ultrasonic cement extractors or hand held instruments.
If the bone – cement interface is poor then rapid progress can be made using hand instruments without damaging the residual femoral cortices. Here shown is a cement splitter for disrupting the bone – cement interface, using the same careful and patient techniques.

Tibial tubercle osteotomy is prepared to facilitate tibial cement extraction. Despite removing the proximal cement from the tibia, distally the cement is well integrated, so the decision was taken to do a tibial tubercle osteotomy (TTO) to provide access to the medullary canal for cement removal. Indications for a tibial tubercle osteotomy in revision knee arthroplasty include: to provide access to the joint e.g. stiffness, explanting the prosthesis and/or cement, proximalising the tuberosity for patella baja, and medialising the tuberosity for patella instability.
To begin this, the skin incision is extended along the medial border of the tibia as shown. I prefer the incision distally to stay medially in case a medial gastrocnemius flap is ever needed in the future for soft-tissue reconstruction. A nice description of this technique can be found here: https://link.springer.com/article/10.1186/1471-2474-9-98

The tibial tubercle osteotomy is marked. I’ve reinserted the tibial prosthesis to appreciate the line of the tibial stem, running in the mid-point of the tibial shaft.
The middle of the tibia, stem is the target for the longitudinal osteotomy.

Drill holes along the marked tibial osteotomy, prior to starting, to prevent propagation of fractures. The holes are drilled approximately 2-3 cm apart, aiming for the tibial prosthesis stem using a 2.5mm cortical drill. Having the prosthesis in-situ gives the surgeon audible feedback when the drill hits the metal implant.

Complete the tibial tubercle osteotomy using multiple broad flat osteotomes. Again ensuring that these are aiming in the direction of the tibial prosthesis stem, sharp blows facilitate completion of the osteotomy between the pre-drilled holes.

Insert three of the broadest flat osteotomes on the set through the osteotomy towards the medullary cavity, tapping each sequentially so that the stress is distributed along the length of the osteotomy so the tubercle remains in one piece.

Lever open the osteotomy using multiple osteotomes.Once all the osteotomes have reached the cavity left by the prosthesis stem, gently elevate them all together. The osteotomy hinges upon the periosteum and fascia of the anterolateral tibia and anterolateral tibial compartment to ensure preserved vascularity to the tibial tubercle for later healing.

Gentle retraction of the osteotomy exposes the medullary cement of the meta-diaphysis (C) for extraction under direct vision.

Tibial cement extraction using hand instruments. The vision and access permitted through the osteotomy permits complete cement extraction along the meta-diaphysis.

Perforate the cement pedestal distally at the isthmus using an ultrasonic probe. The OSCAR (Orthopaedic System for Cemented Arthroplasty Revision, Orthofix) has single use probes receiving ultrasound waves at greater than 16Hz from the handset. It is used to melt the polymethylmethacrelate cement and pierce through the cement pedestal and cement restrictor in the tibial canal distal to the osteotomy.

A flexible canal probe is used to confirm that the host medullary canal has been reached distally after perforating through the cement pedestal using the OSCAR probe, as the temporary spacer will need to be reinforced distal to the isthmus.

Ream the tibial canal at the isthmus using cement reamers. To increase the diameter of the Kuntcher nail (implanted to reinforce the spacer), the isthmus is sequentially expanded using cement reamers until cortical “chatter” is experienced.

Explant the residual patella button prosthesis using the oscillating saw.If the residual patella bone is thin then attempting to remove well fixed polyethylene patella implants risks fracturing and fragmenting the bone with hand instruments.
I prefer to use the oscillating saw to to cut through the pegs by sawing between the prosthesis – cement interface.

Once the three pegs have been divided it should be possible to lift off the patella button without bone loss. This reveals the residual cement and gross osteolysis.

Using a curette, the osteolytic bone is carefully removed and then a high speed burr is used to remove and residual cement in a controlled fashion to avoid fracturing the patella.

After debridement, bleeding host patella bone has been exposed, and unfortunately this patella will be unreconstructable at the second stage due to the poor residual bone stock.

Thorough lavage of all canals and cavities is performed prior to insertion of the cement spacer.Typically during the entire case we would use a 3 litre bag of warmed NaCL 0.9%.

Insert the femoral and tibial Kuntscher nails to reinforce the static antibiotic loaded cement spacer.Usually I would use a Kuntscher nail the same or 1mm smaller diameter than the last canal reamer used using the canal preparation stage. These were the original stainless steel slotted nails first used by Gerhard Kuntcher during WWII for stabilising fractures in German soldiers (before osteosynthesis gained general approval). Alternatives include contemporary titanium medullary nails used in modern practice or carbon-fibre rods used in external fixators, but these Kuntcher nails are cheap and disposable and come in a range of lengths (10cm to 20cm) and diameters (9mm to 16mm).

These Kuntscher nails must be inserted proximally in the femur and distally in the tibia sufficiently to afford stability.
As a rule of thumb if you have a 10cm segmental bone defect you need the total length of reinforcement to be double i.e. 20cm. This means you need at least 5cm of Kuntscher nail in the femur and 5cm of nail in the tibia and 10 cm of nail between the bone ends = 20m total.
However, where you have a tibial osteotomy, the length of the osteotomy should be considered to be included in the segmental defect; thus a 10cm segmental loss plus the 10cm osteotomy becomes 20cm. Therefore you need additional Kuntscher nail reinforcement totalling 10cm in the femur and tibia (distal to the end of the osteotomy).

Once the nails are inserted into the femoral and tibial canals I tap them gently with a mallet to embed them and then use the Frosch bone holding clamp to grab the two nails together.
As Kuntscher nails are slotted, it is advised to ‘slot’ the nails into each other for additional stability. Once clamped you should be able to lift the leg without a ‘lag’ just using the Frosch clamp.

Close the tibial tubercle osteotomy over the Kunscher nail.Using iso-elastic Supercables (Kinamed), I close the tibial tubercle osteotomy by drilling holes using a 2.5mm drill on either side of the osteotomy and passing the cables through the holes and tensioning until bone apposition occurs.
Distally this can be augmented with firewire sutures passed through the cortices on either side of the osteotomy (via holes made with the drill or bone awl).
Alternative methods for closing the tibial tubercle osteotomy include ethibond sutures and partially threaded screws with washers, although some authors prefer not to fix the osteotomy at the first-stage surgery and ‘fix’ the osteotomy at the second stage: https://d1wqtxts1xzle7.cloudfront.net/55076540/59_Tibial_tubercle_osteotomy_in_revision_knee_arthroplasty.pdf?1511319942=&response-content-disposition=inline%3B+filename%3DTibial_Tubercle_Osteotomy_in_Revision_Kn.pdf&Expires=1624202773&Signature=QTCh8mT8wWTH8zerqnzmx1DvA0tP2p~DHYAOjL3DsJwH8trQDdgPTp5O45GVFijcfGvq1ARIV6AH2weWYNWlGmSiMAwRUg1-DWixdz449MjLIlylfC6JfxNRqjER9wjMpW7JNEipqhJf07npy5Yd4EQlOdSW8b~hdhmpNPLAOfSzERC3dGL-jHX6XaU7Mv1XSk4glGkiNh21nmGFHeVj11RlIvq4amNu1ubKDgazbs316clxj~tyHGMMbayptoQRfjDZVUUIqw3T5zqCpLrWZvXBj1YlmfKj74XTwjovrrZDkmegX7OD-DgZodaM4fBBMv0wId8AXvUdv0-DFga~Bw__&Key-Pair-Id=APKAJLOHF5GGSLRBV4ZA

Prior to adding the antibiotic loaded cement, the segmental defect is stabilised and the osteotomy closed.

Prepare the antibiotic loaded cement spacer.Palacos G high viscosity cement is prepared routinely with additional antibiotics added as required which are determined by the pre-operative MDT discussions with infectious diseases and microbiology colleagues.
In this case I used additional vancomycin (2 grams per 40 gram cement mix) as the organism was sensitive to vancomycin. Pump the cement onto the nails whilst the assistant maintain length and rotation of the limb (hopefully alignment is already controlled by the Kuntscher nails). The additional antibiotics means the cement will set slowly, so be patient.

Once the cement has covered the nails and ends of the femur and tibia a tube of cement covering the nails should be established. To increase the surface area for antibiotic delivery the wait for the cement to cure can be filled by using forceps to make imprints in the cement surface as shown.

Additional local antibiotic delivery can be enhanced using Stimulan beads.Calcium sulphate beads (Stimulan, Biocomposites) with additional vancomycin (2 grams per mix) are mixed and set and added to the effective joint space for additional delivery of antibiotics to the joint. The theoretical risk of this is that they cause a greater joint effusion, so a ‘watertight’ closure will be mandatory.

I use a looped 1.0 PDS monofilament suture to close the arthrotomy from distal to proximal in all joints. This is particularly in infected revisions, for a watertight closure of the arthrotomy.
Ensure you bury the knot proximally to avoid local tissue irritation from the knot.

Skin closure using skin clips.

Wound coverage using a vacuum assisted dressing. Although the scientific evidence is equivocal, all high-risk wounds receive incisional vacuum dressings. High-risk in my practice includes previous radiotherapy, infection including PJI, malnutrition, any muscle flap and multiple operations through the same incision as in this case.

AP radiograph of a static reinforced distal femoral spacer showing the medullary Küntscher nails.
These are overlapping at the site of the excised distal femur (A) wrapped in antibiotic loaded cement with skin clips from the previous closure and evidence of a (S) super-cable clip closing the proximal extended tibial osteotomy used for explant at the first-stage.
The patient also has a splint on shown on his radiograph.