Revision Total Knee Replacement- First stage for acute Prosthetic Joint Infection (Zimmer-Biomet articulating spacer)

Pre-operative AP radiograph showing a cemented condylar resurfacing total knee replacement which is cruciate retaining. The clips and surgical drain are commensurate with recent implantation. The patient was diagnosed with endocarditis using echocardiograms with peripheral blood cultures identifying the same organism (& sensitivities) that was obtained from the pre-operative aspiration prior to the first stage revision procedure.

The patient is positioned supine under an appropriate anaesthetic with targeted antibiotics intravenously at induction with a thigh high tourniquet and an exclusion drape applied. The skin clips from the index total knee replacement have just been removed which is why there is some blood on midline the wound.

After double alcoholic Chlorhexidine skin preparation, the leg is draped with a bowel bag over the foot and Ioban incisional drapes as shown. After WHO check and leg elevation, the tourniquet is inflated.

The previous midline skin incision is used; upon reaching the deep fascia the joint cavity is aspirated with a needle and syringe to obtain fluid for further microbiological culture and sensitivity routinely.

The synovial fluid is transferred into aerobic and anaerobic blood culture bottles as shown routinely to optimise culture yield and sent to the laboratory.

After defining the soft-tissue layers using McIndoe dissecting scissors, the previous medial para-patella arthrotomy is used to expose the joint as shown.

The first stage of the explantation is to begin the anterior synovectomy on the medial aspect of the knee to facilitate radical debridement, to provide tissue culture samples and to make additional space in order to facilitate explantation. All debrided tissue is sent for culture (5 samples) and histology (2 samples) routinely at both first and second stages. Lane tissue forceps are carefully placed on the edge of the arthrotomy medially to maintain tension and the inflamed synovial layer (S) is dissected using cutting diathermy from the suprapatellar pouch to the medial gutter of the knee and just proximal to the medial collateral ligament (MCL), in one piece.

Next the anterior synovectomy moves to the lateral side of the joint. (K) Kocher’s and (L) Lane forceps are used to grasp the synovial layer and arthrotomy respectively to maintain tension for debridement from the suprapatellar pouch to the lateral gutter and immediately proximal to the lateral collateral ligament (LCL), again in a single piece of tissue.

The debrided (S) synovial tissue is removed from the medial side of the distal femur using large hand held nibblers which enables the synovial layer to be stripped away from the bone without damaging the collateral ligament insertion.

Clinical photograph showing the (MCL) medial collateral ligament origin on the distal femur has not been molested during the radical debridement using the tearing technique of the (S) synovial layer. The polyethylene of the tibia is shown indicating that the sheet of tissue debridement has been maintained in continuity from suprapatellar pouch to the tibial insertion.

Following thorough anterior synovectomy improved access to the distal femur permits explantation of the condylar femoral prosthesis. A broad, flat (O) osteotome is inserted between the cement-prosthesis interface and tapped gently beneath the (A) anterior flange of the prosthesis. A tip is to keep the osteotome aiming towards the undersurface of the prosthesis, to avoid gouging into the bone which can result in unnecessary bone loss.

Smaller osteotomes are used medially and laterally to loosen the prosthesis at the distal femur as well as the anterior and posterior chamfers. The same techniques to avoid bone loss are used: identifying the cement-prosthesis interfaces and aiming the osteotome towards metal rather than bone.

A hand held Gigli saw is inserted under the anterior flange of the femoral prosthesis and (keeping your arms as wide as possible) the Gigli saw is brought distally towards the surgeon to loosen the cement-prosthesis interface. The knee is positioned at 45 degrees of flexion. This saw not pass the pegs of a cruciate retaining prosthesis, nor the box in a posterior stabilised prosthesis; if you persevere the Gigli saw will snap. Take care to retract soft-tissues during this step to avoid damage as shown.

The prosthesis is gently tapped out using a punch on the tip of the anterior flange. It can be seen that the prosthesis has been removed without significant bone loss on the femoral prosthesis, which is common with a careful explantation technique. In more osteoporotic bone and chronic infection, the bone loss may be more severe.
At the time of reconstruction the options for replacing lost bone include cement (in small defects <5mm), metal augments on the tibia and distal & posterior femoral condyles (>5mm), trabecular metal cones (for deficient epiphyseal bone stock) which are independent of the implant or sleeves (for deficient epiphyseal bone) which are integral to the prosthesis, and morcelised or bulk auto- and allograft. In gross bone loss endoprosthetic replacements may be needed.

Three retractors are used medially, laterally and posteriorly to gain access to the tibial prosthesis. A broad flat (O) osteotome is inserted again between the prosthesis and cement interface to loosen the tibial prosthesis. This is a monobloc tibial prosthesis so the polyethylene cannot be removed separately from the tibial tray. In this implant the polyethylene overlaps the edge of the tray, so be aware when identifying the cement-prosthesis interface.

A lateral view shows the direction of the osteotome being inserted between the cement and the tibial prosthesis. Initially the osteotome is inserted centrally towards the central keel of the prosthesis….

Once the keel has been reached (signified by a change it the note of the hammer on the osteotome) the osteotome is rotated by the surgeon to remain in the same plane between the cement and the prosthesis and to loosen the interface on the medial side of the prosthesis. A tip is to check the distance the osteotome needs to reach the posterior rim but superimposing over the tibial tray before insertion, to avoid the osteotome travelling beyond the bone into vital structures.

Once the interface between the cement and the prosthesis has been broached using the osteotome medially and laterally a double osteotome technique is used to explant the prosthesis. A broader osteotome is inserted on the medial side and then another osteotome is inserted immediately above that and gently tapped in to position to spread the load on the host bone (to avoid tibial bone loss) and to elevate the prosthesis from the cement mantle.

Once loose, a generic tibial prosthesis extractor is used to remove the implant as shown. This extractor has a pair of forceps which are placed around the anteromedial and posterolateral corners of the tibial prosthesis and an integral slap hammer to tap it out. Usually the lateral femoral condyle needs to be retracted to enable this step, so move the posterior tibial retractor more laterally to facilitate this.

Two osteotomes are then used to remove the cemented patella button as shown. Again the osteotomes are inserted into the interface between the cement and the prosthesis to minimise bone loss.

Once the patella prosthesis has been removed, the residual cement is removed using a high speed burr with a 4mm rounded tip.

With the implants now explanted, the residual synovium is radically debrided. On the medial side the McIndoe scissors are used to dissect around the medial side of the joint sparing the (MCL) medial collateral ligament which is highlighted at the tip of the scissors.

A laminar spreader is used to separate the femur and the tibia and to maintain tissue tension whilst the synovium at the back of the knee joint is debrided using cutting diathermy in a single layer from medial to lateral including the femoral notch. All tissues are sent for culture or histological analysis. The neurovascular bundle of the popliteal fossa lies posteromedially and care should be taken to achieve a thorough debridement without damaging these vital structures. The popliteal fossa contains the popliteal artery and vein, tibial and peroneal nerves and short saphenous vein. The deepest structure and at greatest risk during the posterior capsular debridement is the popliteal artery.

Lastly, loose (infected) bone and residual cement is removed from the tibia and the femur using spoons, curettes and nibblers leaving only host bone. Be sure to curette the peg holes and remove all tibial cement from around the keel of the prosthesis.

The knee is now extended and thoroughly washed using copious (3 litre bag) sterile saline 0.9% which is warmed. Aqueous betadine may be added to the saline (30ml per litre concentration).

The tourniquet is then released and haemostasis is secured as shown. Check homeostasis in extension and flexion as different vessels may be identified in different positions.

An incisional vacuum dressing is used on high risk wounds in my practice and this is attached to the knee as shown. The wound is considered high risk because of the patient’s clinical state and because of the recent index procedure.

On the back table the molds for the pre-formed articulating spacer are shown. The (T) tibial prosthesis to the left forms a simple cup which has measurements on it to assess the depth of the final product. The (F) femoral mold on the right has a round aperture for insertion of the cement gun as shown. The molds come in sizes according to anteroposterior and mediolateral dimensions (which can be measured off the explanted prostheses).

A triple mix of Palacos high viscosity cement (with 2grams of Vancomycin and 1gram of Meropenem per 40g mix) is injected using the nozzle in to the femoral mold as shown.

One tip to enable the cement to evenly spread around the mold is to take a needle and insert it in to the corners as the mold is being filled to allow the air to escape from the mold, as shown. Do not over fill the mold.

The tibial mold is then filled to the desired depth with the same cement and allowed to cure.

Once the cement has hardened, a scalpel is used to incise the femoral mold down the centre…

The (F) femoral mold is then gently split by the surgeon to gain access to the spacer. The (T) tibial spacer is also shown on the right hand side of this clinical photograph. It can be seen that the preformed tibial mold incorporates a lip to give it some anteroposterior stability.

The articulating spacers are now ready for implantation. The tourniquet need not be re-inflated as cement pressurisation is not desired.

The (T) tibial spacer is inserted first, my preferred technique is to fabricate a cement keel to fit in to the proximal tibia to prevent the tibial spacer from dislocating using a separate mix of cement (with antibiotics added at the same concentration) and then attached to the under surface of the tibial articulating spacer. The (F) femoral spacer is ready to implant.

The femoral spacer is then inserted with the knee in flexion as shown and held in place by the surgeon. It should achieve a snug fit and require significant pressure to ease onto the bone.

The knee is now extended (carefully) with the spacer in-situ and a check is made to ensure that the femoral spacer does not subluxate in to extension. Carefully, collateral stability is checked as gross joint instability is a contraindication to an articulating spacer. The joint cavity is now thoroughly washed.

Closure of the tissue layers, starting with the arthrotomy, begins using a looped monofilament PDS suture No.1. No drain is inserted to maximise the concentration of any antibiotics that leak out of the articulating spacer.

Soft tissue layers are closed routinely using interrupted Vicryl No. 1.0.

The skin is then closed with clips as shown.

Immediate post-operative AP radiograph showing the cement spacer in-situ. Surgical clips are shown. The additional keel attached to the under surface of the tibial spacer is shown.

A lateral post-operative radiograph again showing the articulating spacer in-situ with the keel and showing no immediate complications. This patient was able to walk without crutches on this spacer between his staged revision procedures.

Post-operative AP radiograph after the second stage showing a fully cemented semi-constrained stemmed knee prosthesis (a Zimmer Biomet DA 360 SSK). Stemmed implants are recommended in the revision setting with this prosthesis, the standard length is 80mm but shorter (40mm) and longer (up to 300mm) are available.

Post-operative lateral radiograph following second stage semi-constrained revision-type prosthesis.