Revision total Knee Replacement- Legion CCK (Smith and Nephew)

Reasonable overall alignment in terms of weight bearing axis, but with slight valgus of the tibial component. However the most obvious abnormality is wear on the left TKR polyethylene with reduced thickness medially.

Significant wear of polyethylene in this Kinemax TKR. No gross lucency in the tibia or femur.

Reasonable congruence on skyline with and un-resurfaced patella.

CT scan showing good rotational profiling with the femoral component slightly externally rotated from the trans-epicondylar axis which runs form the sulcus of the medial epicondyle to the ridge of the lateral epicondyle. The posterior condylar line of the implant of the implant (external geometry) should be slightly (ideally 3 degrees) externally rotated from the transepicondylar line.

Sagittal CT showing some possible lysis in the tibia and concern over Poly wear.

Coronal CT showing significant medial polyethylene wear and lysis in the tibia.

The planned incision (usually the most recent) is then excised and the surgeon’s chosen approach made to the knee. This should be the standard approach used by the surgeon, for most this will be medial parapatella, but for some surgeons a medial subvastus will be the preferred approach for primary TKR; I have used the subvastus in revision TKR for nearly 10 years now without regret. However it is important to know how to extend your chosen approach to relieve tension to reduce the risk of extensor rupture and to provide better access.
In this slide the Vastus Medialis is labelled 2 and the patella is labelled 3.
Proximally this is usually by a mid quads slide (an osteotome is useful for this) or a quads snip, very rarely needing a full quads turndown; the proximal extension options are not available to subvastus users. Distally however this is easily achieved with a tibial tuberosity osteotomy; this provides the added advantage of potentially changing the patella height in cases of Baja.
Irrespective of chosen approach it is vital to raise thick flaps to minimise vascular problems with the flap. – see image marked 1.
The next stage will be the start of sampling with a fluid sample through the capsule not through skin, but in order to keep clarity of writing, I will deal with all sampling now. We follow the Oxford sampling advice (Atkins BL, Athanasou N, Deeks JJ, Crook DW, Simpson H, Peto TE, McLardy-Smith P, Berendt AR. J Clin Microbiol. 1998 Oct;36(10):2932-9.) and perform sampling under the polyethylene, deep and superficial samples from tibia and femur, plus or minus the patella depending on patella management at the primary and the cause for revision; a well fixed patella component should not be removed unless a diagnosis of infection is very likely or known.
The fluid should be tested in the Operating Room/Theatre for the presence of white blood cells on leukocyte esterase strips (‘urinalysis dip sticks’), with 2+ or 3+ being consistent with a very high chance of infection. Alpha Defensin testing of synovial fluid is another useful test to diagnose infection.
In addition a good section of synovium should be sent for histology.
Synovectomy. A formal synovectomy from all around the knee should be performed. This really helps with access, and is vital for the management of infection; unfortunately some diagnoses of infection are only made after the revision specimens are returned, so it is important to perform a thorpugh debridment in all cases. This synovectomy will recreate the medial and lateral gutters and supra-patella pouch.
Implant interfaces. These should be carefully exposed such that the correct interface can then be addressed during extraction.

I use the medial subvastus approach, but I would advise using your standard approach and for the majority of surgeons this will be the medial parapatella. This should be performed in your standard way, but with the extra care and precaution that a revision procedure demands.
Rectus is labelled 1
Vastus Medialis is labelled 2
The Patella is labelled 3

Usually the first step of the extraction phase is removal of polyethylene. Knowledge of the implant being removed is essential as this may require fixation clips to be removed first or particular ‘tricks’ as to the angle of insertion of the operators favourite removal tool. Remember some implants are non-modular monobloc constructions of moulded polyethylene on a metal baseplate giving the impression of modularity on the X-Ray but behaving as an all polyethylene construction on the tibia; these need to be removed in one from the tibial bone-implant interface. Another important ‘must remember’ point is to check that the poly already inserted does not have a stabilising post or locking mechanism down into the tibial metal baseplate.
My favourite extraction tool is the Bristow as it has a sufficiently sharp tip to enter the polyethylene, yet is robust enough to lever on the metal tibial baseplate to lever up the polyethylene.
In this image the Bristow is being inserted between the poly and the tibial baseplate.

After removal of the polyethylene a sample is taken from the ‘hidden’ surface of the tibial component with Oxford sampling technique (5 blade handles and 5 forceps) to minimise cross-contamination from samples.

The bone-component interfaces are then cleared with sharp dissection.
There are various ways to remove an implant – narrow saw blades, plaster saw, osteotomes or Gigli saw. My preferred method is the double impaction technique where a shock-wave is created by simultaneous osteotome strikes from different directions at the implant-cement interface – most commonly medial and lateral at the same time; the best practical advice on this is to insert the osteotome on the medial (assistant’s) side and ask the assistant to strike this osteotome with a regular timing. The surgeon the inserts and osteotome (1) from the lateral side and simply strikes ‘on the beat’ with the assistant’s rhythm!

In uncemented implants the double impaction technique is less rewarding and I personally use a combination of saw and osteotomes.
After the component has debonded it can be removed easily by hand, but in cases where more force is required, extraction sets may be used, for example the Smith and Nephew Renovation set which has a very good slap clasping device which attached to a slap hammer.

The femoral component has been removed and attention turned to the tibia where an osteotome is inserted under the tibial component, above the cement mantle. The alternative technique of fine saw on the cement interface is demonstrated below.

Note the swabs protecting the soft tissues. As the saw generates cement debris this is simply collected on the swabs which are exchanged as the tibia component is removed.

After removal of the tibial component it is important to take out any cement plug within the tibial canal. This can easily be reached with standard osteotomes and then simply pulled out with a clamp such as a Kocher.

Once the components have been removed. Specimens are taken (superficial and deep tibia and femur) then pulsed lavage should be performed. I prefer to use aqueous chlorhexidine for the pulsed lavage irrigation solution. This can be seen as the frothy fluid in the centre of the tibia.
The extraction, sampling and cleaning is complete and we are now ready for reconstruction.

The extraction is now complete. Notice the broken polyethylene and delamination medially and laterally.
Please refer to the future Principles of revision section for step by step guides for extraction, but the above slides are a summary.
Standard implants generally have small cement mantles easily removed with osteotomes, but beware fully cemented long-stemmed implants, these may require Ultrasonic cement removal equipment analagous to revision hip surgery. Similarly combined suction and light devices such as Vital View TM can be really helpful. Where a long tibial component is well fixed in a mantle of cement, it is worth considering a TTO, not just for improved access, but also for allowing acess to the mantle from within the tibia, allowing easier removal and reduced risk of tibial fracture.

Once extracted, currettage deep in the femur and tibia, pulsed lavage with chlorhexidine and then repeating of this whole cycle should be done until the surgeon is confident of a achieving a complete extraction.
My personal practise is to place a chlorhexidine soaked pack in the intramedullary canals, change gloves and then start on the reconstruction.
In this image the swabs have just been removed to demonstrate appropriate access and a clean Tibia, Femur and Patella ready fro reconstruction with the Legion revision system.

The tibia canal is reamed with sequential reamers. I prefer to perform this by hand rather than on power, although the reamers can attach to a power driver.
The reamers go up in 1mm increments from 9 to 16 and then 2mm increments from 16 to 24.
There are various stem lengths: 90, 120, 160 and 220mm. In order to unify the system (using the same stems for the tibia and femur) there is a key measurement difference between the tibial and femoral preparation. The tibial preparation assumes that an offset will be used and the measurement of the construct length comes for the cut bone surface.
This is in contrast to the femoral preparation which assumes that there is no offset used and the measurement of stem length is from the joint line – ie the most distal part of the femoral component.
For most medium sized patients the most common stem length is 120mm, with width being determined by achieving some chatter with the cortex. On the femoral side 160mm is the most common length.

It is sensible to counterbore over the initial reamer, to save time, although in this image the stem extender and trial stem is being used..As you will appreciate this is a large reamer and serious caution is needed, particularly in view of the large lever arm holding the driver.
The counterbore creates a wider canal to sit the tibial boss. If the initial reamer is 18mm or larger then no counterbore is required

The counterbore is inserted on power down to the ‘stopper’ set on ‘tibia’. Again BE CAREFUL!

This image shows a ‘120mm’ stem with an offset on the definitive implant. This means that the distance from the tip of the stem distally to the bone side of the tibial component is 120mm

The tibial cutting block is then applied either directly to the reamer or to a trial stem with and additional stem extension.
There is a 1mm stylus allowing a minimal freshen up cut. The cutting block is a zero slope block so there is no need to pin this in situ and the block can be secured with screwdriver tightening alone, although pin holes exist should the surgeon prefer as in the next image. The advantage of avoiding pinning means the block can be rotated around the reamer or stem extender allowing easier access for the proximal saw cut of the tibia.

Pins can be used to secure the tibial block. The pin holes can be pre-drilled depending on surgeon preference.

Appropriate care should be taken with the saw during proximal tibial cut, particularly with the Patella Tendon, Medial Collateral Ligament and of course caution and protection posteriorly for the neurovascular bundle

The proximal tibia is sized and an estimate made of the difference between the centre of the cut-out in the sizing template and the stem extender. This is the difference between the centre of the metaphysics and the centre of the diaphysis. This difference gives rise to the concept of ‘offset’. There is considerable anatomic variation in metaphysical – diaphyseal offset with the mean offset being 4mm, but a selection of offset couplers is available on the Legion system of 2, 4 and 6mm.

A 6mm offset was most appropriate in this case to match the black tibial sizing template with the stem extender on the trial stem in the diaphysis of the tibia. The offset coupler is rotated until the best fit coverage and rotation of the tibial component is achieved.
We usually record this as a 6mm offset at 5 o’clock, describing the number on the tibial sizing template that the 6mm offset coupler’s arrow points.

The tibial baseplate template is then pinned in position. There are four holes available in the template, but beware situations when there is a defect medially or laterally as the template will not sit down on the bone over such a defect; the medial or lateral augments are cut in at a later date. Consequently the pins should be inserted into good bone stock first.

The stem extender and offset coupler have been removed and the central hole in the templates been filled with the centralising collet for the tibial reamer to allow the tibial boss to offset in the same position as the template.

The tibial counterbore reamer is then inserted, down to the positive stopper set on ‘tibia’.

The tibial trial is then prepared with the appropriate sized tibial baseplate, offset and stem. Note that the stem and the offset attach via a spring-loaded bayonet-type fixation. This is important as the trial may be very slightly longer than the length of the prepared bone by the spring loading distance. The next slide shows how to address this.
The counterbore reamer is used twice. The first reaming is over the stem extender or original hand reamer to widen a sufficient margin to accommodate the offset coupler (annotation 2). The 2nd counterbore reamer use is to accommodate the tibial boss (annotation 1).

If the tibial baseplate ‘sits up’ on the cut tibial bone, the stem needs to be inserted slightly deeper to accommodate the spring-loading of the trial. This may be achieved by a gentle tap of the Legion Set Hex Screw-Driver located in the grub screw situated in the offset coupler, accessed through the centre of the tibial trial plate

By attaching drop rods (ideally 2) then the alignment of the tibial component can be checked in the sagittal, coronal and axial planes, before this is confirmed by preparing the tibial keel or any augments that may be needed. Should an agent be needed then this is cut using a block attached to the quick connect handle pictured here, clipped in to the tibial trial and holding the 2 drop rods.

Attention is then turned to the femur and in a similar manner to the tibial preparation, the intra-medullary canal is reamed until good interference fit is felt. As in the tibia I prefer to use hand reaming.. Care should be made when introducing the reamers, by comparing the diaphyseal alignment and position relative to the metaphysics on the long-leg radiograph displayed in theatre.

The distal femoral cutting block is attached to the reamer or a trial stem and stem extender as an alternative.
I prefer to work out joint line position using the Romero method of multiplying the transeipcondylar width by 0.4. This gives the distance form the medial epicondylar prominence to the joint line. By subtracting the 10mm of the femoral component this gives the distance where the distal femur should be cut – of course if there is bone loss then this will need building up, but this is a useful tip to help determine a good joint line position.
The transepicondylar width is measured simply by a ruler.

Be sure to check that the 6 degrees valgus central bushing is inserted with the ‘lateral’ marking on the lateral side!
Here the angel wing is placed through the distal cutting slot on the distal femoral cutting guide. A very useful trick is to use the Romero joint line calculation mentioned earlier and to work out the distance from the medial epicondylar prominence to the joint line and the back of the femoral component…..this allows an estimate of appropriate augment use before the cuts are made.
For Example if the trans-epicondylar width is 80mm in a large patient then the distance from the joint line to the medial epicondyle will be 0.4 x 80mm = 32mm. As the implant is 10mm then the distal most bone should be 22mm from the medial epicondyle. Similarly, in a small patient with a width of 62mm the the joint line should be (0.4 x 62mm) 24.8mm from the medial epicondyle. 25mm is often quoted as the distance for the medial epicondyle to the joint line, but really this depends from where on the epicondyle the measurement is taken!

The block is pinned (drilled and pinned ideally) and then the distal resection is made, sometimes allowing for a distal femoral augment and cutting through the relevant slot (5,10 or 15mm).

Next the extension space gap is checked. Here the 9mm black gap measurer has an additional 5mm augment laterally on the femur (yellow) and 5mm laterally on the tibia (purple).

Full extension is achieved with the spacer block and there is good medial and lateral balance of collaterals in extension. The block allows for drop rods to measure alignment too.

With the femoral trial stem inserted on a stem extender, an appropriate size femoral sizer is then used to assess how much offset is required anteroposterior and how much augment will be needed posteriorly (5 or 10mm are available).

In general posterior offset is the normal finding and we describe a pure 4mm posterior offset as “4mm at 6 o’clock”, but if there was a need to offset medially as well then a 6mm offset could be taken but positioned at “7.30” to allow some of the 6mm to offset posteriorly and some of the offset to correct the femoral metaphyseal-diaphyseal mismatch in the medial lateral direction.

If the stem lines up with the 0mm then no offset is required on the femur. However this is very unusual and most cases require a 4 or 6mm offset in the posterior direction somewhere between 4 and 8 0’clock.

The femoral cutting block turned around to show the distal femoral augments on both sides. (Medial = 1, Lateral = 2)

Adjusting the central offset fine tunes the position of the chosen femoral cutting block relative to the stem and extender. Here the offset is positioned to 3 o’clock which on a left knee would offset the cutting block and ultimately the femoral component to the lateral side.

The offset has been rotated around to 6 o’clock, pushing the cutting block maximally posterior, filling the flexion space and thus reducing flexion instability. Caution is needed not to notch anteriorly; if this is a risk then the offset amount needs to be reduced or the femoral component upsized.

Femoral cutting block in position with offset at 6 0’clock (pointing maximally posteriorly).

With the femoral block in position, an angel wing is used to check there is sufficient bone posteriorly. This is an open cutting block, with two slots at 5 and 10mm, as demonstrated on the lateral side with this angel wing here.

After checking with the angel wing in last slide, the offset (1) is fine tuned to 5 o’clock, moving the block slightly lateral and anterior from the 6 o’clock position. This allowed sufficient bone contact posteriorly in this case using the maximum 10mm posterior augment on the lateral side.
The block is pinned in place and the cuts made.

The stem, extender and offset are removed and a centralising bushing placed in the cutting block. The large femoral counter bore is then used similar to on the tibia, reaming 18mm out for the femoral boss and offset.

The femoral stem and extender is then replaced and the counter bore (set on ‘femur’) is then used over the stem extender to ream out sufficient space for the offset coupler. Remember that if an 18 mm stem or greater was used earlier then there is sufficient diameter not to perform this step.

The femoral trial is prepared with augments distally and posteriorly by using an ‘L-wedge’ 10mm x10mm (1). The offset coupler is labelled 2 and the stem 3.

Once the trial is assembled, the position of the offset coupler is adjusted to dial in the required position as described by the 5 0’clock position on the femoral cutting block and offset earlier.
A useful operative tip here is to support the hex head screwdriver (1) on the surgeon’s abdomen, whilst the surgeons left and right hands control the femoral component and stem/offset combination separately. Once the stem/offset coupler is at the required position (2) then the screw driver (1) should be used to tighten the central grub screw inside the coupler.

The femoral component is introduced into the femur and with the surgeon controlling the component rotation to match up with the cuts, an assistant surgeon impact the implant trial down to the cut femoral surfaces.

With the femoral trial in position, the box cutting guide (1) is inserted.

The femoral box is then prepared by reaming in the anterior and posterior position that the guide will allow by sliding anterograde-posteriorly.

The final preparation of the box is with a specific box chisel in both the anterior and posterior position of the box guide. Here the anterior position is found by using the surgeon’s thumb to push the guide anteriorly.

The box cutting guide is then moved posteriorly and the chisel repeated.

The femoral box housing unit is fitted – this accepts the polyethylene post either in condylar constrained or posterior stabilised insert varieties.

The leg goes into full extension at the knee with the trial in situ.

A good flexion range is achieved – approximately 11o degrees here before posterior soft tissue impingement. Patella tracking is checked with the native patella in this case or with a the previously replaced patella in most revision TKRs, excepting infected cases where all implants need revision.

The position of the proposed joint line (femoral component to polyethylene bearing interface) is checked relative to the meniscal scar (1) which is highlighted by Hohman retractors marking the superior and inferior surface of the meniscal scar on the medial capsule.

The patella is prepared in a standard Genesis II fashion. I always resurface the patella at revision TKR if not already resurfaced. For cases where the patella is resurfaced I assess the state of the polyethylene, angle of the patella cut and implantation as well as the width: thickness ratio. I use the patella calliper to measure Bothe width (as in this photo) and then thickness.
Bristol index of patellar width to thickness (BIPWiT): a reproducible measure of patellar thickness from adult MRI.
Knee.2014 Dec;21(6):1058-62. doi: 10.1016/j.knee.2014.07.007.
Sullivan NP1, Robinson PW2, Ansari A2, Hassaballa M2, Robinson JR2, Porteous AJ2, Eldridge JD2, Murray JR2.
The Genesis has two options: onlay or inlay.
I prefer the inlay single peg biconvex patella as my standard patella replacement button in primary TKR and patellofemoral replacement cases, so I also use this in revision TKR for most cases.

Measuring patella thickness.
I aim for an overall width/thickness approaching 2:1. From the study referenced above we found a fairly constant ratio of 1.8 in healthy patellae on MRI.

The patella is reamed sufficiently to reconstruct to the desired thickness including the patella button.

The implants are opened in modular form and assembled to match the trials and then impacted.
It is a good idea to open the tibial and femoral components separately so that there is no chance of mixing up offsets or stems.

The packaging around the implants forms an excellent protection for the implant during impaction. One single hit will impact the morse taper perfectly well.

There are 4 grub screws – 2 for the offset and 2 for the stem.

For the femoral preparation the distal augments are secured first – I find the standard Hex screwdriver is quickest and then using the torque-listed T handle for the last couple of turns.

Torque-limited T-handle for the final tightening of the augments.

The offset and stem are added in the correct position.

The stem/offset combination is then impacted onto the femoral component.

A hybrid cementation technique is my standard. I use Palacos regular cement with Gentamicin for standard cases. This is gunned into the cut tibial surface, but with care not to cement the stem.

The back surface of the tibial component and offset are coated with cement, prior to implantation.

The tibial component is impacted and then excess cement is removed.

A similar technique is then used for femoral implantation, with a fresh mix of cement (mixed as I take the ‘tibial’ cement) gunned onto the distal femur and the implant.

Cementation on the femoral implant and offset region – effectively zones 1 (cut surface of the femur or tibia in a primary TKR) and 2 (metaphysis)
Zonal fixation in revision total knee arthroplasty.
Morgan-Jones R, Oussedik SI, Graichen H, Haddad FS.
Bone Joint J. 2015 Feb;97-B(2):147-9. doi: 10.1302/0301-620X.97B2.34144.
PMID: 25628273

Femoral component being inserted.

Impaction of the femoral component.

The trial polyethylene is inserted and the leg extended to compress the cement interfaces.
At this stage the patella would be implanted as in a primary TKR.
An assessment of bearing size is then confirmed as well as the stability of the knee. In general I will try to insert a posteriorly stabilised insert in most Legion Revisions, but if there is concern over laxity then a condylar constrained knee (CCK) insert should be used. In this case a CCK insert is in position

The definitive CCK Legion bearing is inserted and clicked into position as the leg is moved into extension.

The medial subvastus approach is repaired – I favour a heavy barbed suture such as 2 Quille or 1 Stratafix, starting at the apex in the centre of the incision (superomedially to the patella) and suturing both proximally and distally.

After the capsular layer is closed, I use interrupted 0-Vicryl in the deep dermal layer, with the knee in flexion.

Interrupted sutures have been placed so I then use running 2-0 deep dermal and a 2-0 or 3-0 barbed subcuticular suture such as Quille or Stratafix, again starting centrally and running distally and proximally.Again I like to suture the knee in flexion.

Skin glue and obliquely placed suture strips – ideally original Steristrips are then used, prior to a clear occlusive padded dressing such as LeukoMed, Tegaderm Pad, Opsite Postop, Hydrofilm plus etc.

Postoperative radiographs – AP and Lateral should be taken as soon as possible. These were done the morning after surgery.
The ideal postoperative image will show appropriate sized implants (matching the original native knee anatomy), no over hang of components, good interference of the stems in the diaphysis, no excess cement.

Lateral day 1 radiograph.
In addition to the ideal findings above on the lateral radiograph, there should be no cortical breaches, good restoration of the posterior condylar offset ratio (PCOR) and minimising the anterior trochlea offset – ie the implant should be flush with the femoral anterior cortex.
The PCOR should be 40-50% – ie the ‘height’ of the posterior condyle on the lateral XR (posterior to the posterior femoral cortical line to the posterior most part of the femoral implant) divided by the total anteroposterior width of the femoral component.