Ankle replacement-Revision using Wright Invision Ankle replacement system

The pre-operative plain X-ray, hinting at a significant cyst associated with the talar component.

The pre-operative coronal CT confirming the extensive nature of the talar cyst.

The saggital slices in keeping with the coronals.

The sterile Invision Jig was built on a side table according to the manufacturers instructions to facilitate accurate positioning of the limb with 2 planar orthogonal image intensifier guidance (1). The jig needs to be carefully constructed and the alignment of the guides will have already been checked by the manufacturers team in the preceding few days prior to sterilisation.
The jig is built whilst the patient is being anaesthetised and takes around 20 minutes.
It is my practice to ensure that all the equipment is sterile and the wraps have not been damaged prior to the anaesthetist starting any invasive procedures as sometimes the drapes can have tears in them thus we are unable to confirm the kit is sterile and the procedure has been abandoned.

The anterior cutting jig is also assembled

The previous incision is marked and subsequently used.

In every case the leg is prepped to the knee to allow accurate on-table alignment of the prosthesis.

I prefer to operate on cases such as this which may take a long time without a tournquet inflated. The thigh tourniquet is applied in case it needs to be inflated for parts of the procedure which in my experience is more often the start of the operation, when the old implant is being located and scarring is encountered. In this case the tourniquet was never inflated.
Great care is needed with the soft tissues as the anterior neurovascular bundle is often buried in scar tissue. Self retaining retractors are to be avoided throughout due the risk of causing damage to the skin edges and subsequent impaired wound healing.

Careful dissection of the anterior capsule to reveal the polythene bearing which is removed whole with Kochers.
Infection should always be considered and blood tests done pre-operatively. However even in the presence of normal blood inflammatory markers indolent infection may be present and only deep specimens will be able to exclude this. A total of 5 samples are sent for culture and sensitivity. I take one from the anterior capsule and one bone sample from each of the metallic/bone interfaces (ie tibia and talus) once theses components have been removed. Two further soft tissue samples are taken from each side of the joint. 5 are used so that there is never an even mix of ‘growth’ and ‘no growth’ samples there by confirming or refuting infection is clearer. This is the same technique used by hip and knee revision surgeons. The bone and soft tissue can look remarkably normal when low virulence organisms are present.

Bony growth around the talus is removed with osteotomes especially laterally. Care is taken not to damage the medial talus wall around the cyst. The cyst is carefully found before the talar component is removed so as to avoid the risk of fracturing the talus if the metallic component needs to be levered out if there has been bone ongrowth onto its backside.

The talar component is removed using osteotomes to free it up from the underlying bone. A knowledge of the undersurface shape of the talar component is really helpful to allow careful gradual bony dissection to get the component out. I don’t want to remove more bone that needed but equally I do not want to have to remove the component with too much force and thereby risk pulling out lots of native talus bone or risking a fracture. In this case the talus came out moderately easily once the undersurface had been mostly freed up. Inspection of the backside reveals some bony ingrowth.

Specimens of the interface are also now sent for culture. A total of five specimens should be sent to confirm or refute infection.

The tibial component which was loose is tapped into the void created by removal of the talar component and the meniscal bearing and an anterior tibial window is then created to allow for the stem to be removed. By tapping the component distally this window size is minimised thus preserving bone stock. It is common on the Mobility ankle system in my opinion to find the tibial component just has membrane grown onto it without evidence of bony ongowth. This may account for the high incidence of medial pain but does make explantation easier. The bone piece used to create the anterior tibial window is kept and may well be put back after the revision has been implanted. I ask the scrub nurse to keep all bone removed in case it is required later in the procedure for grafting the cysts, to provide support as in this case to the tibial window or if all else fails and the procedure is abandoned in favour of a fusion then all bone will be needed to fill the subsequent void.

The tibial component is removed and inspected – little bone ongrowth is seen. The final bone and soft tissue specimens are sent, from the tibial component-bone interface for microbiological culture.

On exploration of the talus cyst and removal of the membrane (A) from within it – the weakness of the talar neck is revealed. Only a shell of medial and inferior talus remains. The talus integrity needs to be enhanced. I think that this loss of integrity is why this patient started to get pain in what had previously been an asymptomatic cyst. In effect he was becoming unstable.

The talus therefore needs to be stabilised and so I plan for this to be done with two Accutrak screws through the head of the talus which is easily visualised. Thought needs to be made about ensuring that the fixation is not too proximal / superficial in the talus to allow subsequent talar plate / implant positioning. The medial cyst is clearly visualised to ensure good position of the Cannulated screws across this void thereby providing further stability.
The small fenestration in the anterior tibial cortex made earlier and used to remove the tibial component can be clearly seen here – A

Lateral image intensifier view shows adequate position of the guide wires. Length is assessed using the extraction measuring device for the screws ensuring the maximum length is selected to provide for maximum talar stability.

The first Accutrak screw is introduced after drilling and countersinking appropriately, the screw can be seen to cross the cyst (A).

The foot plate part of the jig is applied with forefoot blocks adjusted to control forefoot abduction and adduction and two wires into either side of the heel through the jigs ‘heel cups’. These wires are cut short.

The foot plate is now put into the pre-prepared jig and secured with the threaded knobs. The Achilles support is loosened off to allow the limb to sit at 90 degrees at the ‘ankle joint’. The heel can be seen wired to the calcaneus at A

The Achilles support is brought up to the posterior aspect to support the lower leg in the jig. This is to ensure that the tibia does not ‘sag’ relative to the ankle ie that the ankle remains aligned between the tibia and the talus in the sagittal plane. It can be secured with further wires in the tibia proximally. In this case it sat well and securely without this additional step.

Position of the AP and lateral rods is checked with the image intensifier to ensure the ‘gunshot’ sights are perfectly aligned. The jig can be raised or lowered on either side with swabs under it until these views are perfect on the image intensifier. This ensures that all is aligned in the coronal and sagittal planes. The rods must pass up the virtual ‘center’ of the ankle joint and proceed up the center of the tibia on both coronal and sagittal views with the image intensifier. This ensures that when the drill is introduced that it too will be centre on all views. Gun sight views are used to check against parallax.

The ‘gun sights’ can be seen to be incorrectly aligned here in another example case which is a revision of arthrodesis to ankle replacement

……but a small degree of rotation now aligns the sights so that we are safe to proceed.

The hole in the foot plate is marked and an incision with a 15 blade to allow the rod to be positioned and subsequently checked in alignment with the image intensifier. The incision is propagated with scissors to ensure the plantar nerves are swept out of the wound.

Drill is inserted under power using image intensifier to check alignment. This is slowly done in a ‘pecking’ manner to make sure that once the tibia is enetered that that the drill passes centrally and is not directed by any pedistal that may exist around the old stem site. Continuous checking with the image intensifier to make sure this drill passess up the center of the tibia in coronal and saggital planes.

The drill also passes centrally up the tibia – again in this example case.

The pre-assembled anterior cutting block jig is added to the frame.

The anterior jig is brought ‘down’ as close to the ankle as the soft tissues will allow and care taken to ensure that the letters on each side are the same – ie that it is correctly aligned in the frame and not ‘twisted’ down either medially or laterally.

Prior to securing the cutting block, AP and then lateral images are taken to check size and further confirm alignment. The close distal cut to the Accutrak screws in the talus can be seen. The size is determined both on the coronal view as seen here resecting a small amount of tibia but ensuring boney cuts are made but also on the sagittal view as seen next.
The jig is seen to lie perfectly centred relative to the drill tip A and the drill shaft at B. The fins C represent where the new joint line will be.

On the lateral view, saw blades are positioned into the cutting jig to ensure the distal and proximal cuts are correct in terms of minimising bone resection but allowing a sufficient resection for bone growth onto the new implant. Again the close cut to the screws is clearly seen.

The anterior cutting block is secured to the tibia and talus with guide wires. These are introduced under power but with care not to drive them in too far – the image intensifier on lateral view is useful here.

Once secured, the anterior cutting block guides the captive saw to make accurate cuts to the tibia and the talus superiorly. Care is taken to ensure the saw cuts are neither too deep , putting the posterior tendons and neurovascular structures at risk especially medially, nor too shallow to provide for adequate good resections.

Subsequently the anterior cutting jig is removed without disassembling the locking in case it needs to be reapplied for further cuts. The wires put in earlier remain in the cut areas of the tibia and the talus to help view where these cuts have been made.

the set has a ‘corner’ chisel on it which is now employed to clean up the corners of the resection and carefully the boney pieces are removed. Careful inspection of the posterior stuctures and to ensure the posterior part of the ankle is fully debrided of bone.

I use pointed ended ronguers to clear away all scar and capsule posteriorly also.

The lateral image intensifier view shows clear resection levels with no bone left posteriorly and adequate though small level of bone on the talus to accept the Invision Plate

The reaming rod is now introduced and sequential reaming made up the shaft of the tibia. We had sized the implant and the reamer is guided up to ensure that the metaphysis only is reamed – over reaming will render the stem potentially unstable. The reaming rod has a threaded end onto which the cutting heads are attached. These are sized in diameter to ream for the base plate stem and any additional stems required.

The reamer head is held on a specific ‘clip’ and then threaded onto the reamer rod with the rod being turned by hand. The reamer should always be driven clockwise both reaming in and out, to avoid the reamer head unscrewing in the tibia.

Once reamed up the tibia trial is used to ensure that the tibial resection is adequate and to ensure that the cut sized implant is correct on the image intensifier in terms of coverage of the dsital tibia on the coronal and sagittal images.

The trial bearing is put into the tibial component whist stabilising the tibial trail. Then the talus and plate are also now trialled. The plate is introduced and inspected to ensure that this does cover the resected talus surface without excessive overhang. The position of any augments is determined also with the implant trials. In this case the augments would be posterior to where the cyst in fact was and so would add no additional benefit and were not used. This was as was expected on the pre-operative CT scans which showed the cyst anteriorly only.

The genuine tibial component is selected – the stems are introduced on the side table and the component gently impacted into place. Image intensifier is used to ensure it is well seated and aligned.
In this view the screw across the ’empty’ cyst can be clearly visualised A

Trial plate and talus in situ with the genuine tibial component, showing the length of plate needed to cover the cyst. The tibial component can be seen to cover both anterior and posterior tibial cortices too – this is necessary for component stability

Once the plate position has been determined a wire is positioned in the anterior plate trial and the trial removed. This wire determines the plate position. At this stage and before the genuine plate is now introduced the cyst is grafted. Morcellised bone from the resected tibia is impacted into the cyst and mixed with blood this impacted graft in the cyst can be seen on this image (A).

The genuine plate is positioned as can be seen with reference to the wire. Fixing points are now used to hold it in place. The most distal screw can be seen here. The male part of the morse taper on the plate can be seen (A).

Only the anterior fixation can be reliably used as the posterior screws may enter into the posterior facet of the subtalar joint (A). In my experience great care must be taken to screen with the image intensifier in the lateral view to avoid this. In this case only one posterior screw could be safely used.

The talar component is positioned carefully to ensure the morse taper will engage…..

…..and the talar component is then impacted into place with several firm strikes using the specific impactor. It can seen to rest in close apposition to the plate once it is down.

The genuine ultra-high molecular weight polyethylene (UHMWPE) bearing reflecting the size that had been trialled is now ‘thumbed’ into place. In the case of the InBone and the Invision systems the bearing is a fixed bearing as opposed to a mobile bearing. Although there is an instrument for introducing this in my opinion it is cumbersome and no safer than my thumbs for this purpose.

Final introduction is with careful and gentle taps on the instrument designed for this purpose. Care must be taken not to dislodge the tibial component at this stage, now the tibial component is stabilised with the screw in rod anteriorly (A). Great force is not required to seat the mensicus into place.

All three components are in place. Unfortunately at this stage I was aware on close inspection that there had been an intra-operative fracture of the medial malleolus. This not rare especially in revision situations and in those where there was already a cyst in the medial malleolus. This must be addressed however to avoid instability or risk malalignment as we know that addressing this complication intra-operatively reduces the risk of late implant failure according to Gadd, Barwick et al. (see results papers).

In the presence of a stem I prefer to use a locking plate (A), with or without a supplemental screw. The positioning of these is relatively easy under direct vision and with the assistance of the image intensifier. I use a small Zimmer Biomet Midfoot locking plate which offers very good stability.

On this view the locking plate (A) can be seen along with a supplemental screw (B). In fact one of the drill bits broke and so also is seen – clearly a long and complex case!

The lateral view show the implant and medial fixation to be in a satisfactory position. Note the position of the screws in the talar plate also.

Careful layered closure with attention to the retinaculum which I close with interupted vicryl ‘safety’ sutures – ie not a continuous closure in case this gives way. Skin is closed with nylon interupted sutures with minimum tension.
I dress the wound with vasoline gauze, and orthogauze. A well padded backslab is applied for two weeks.

AP view showing components well aligned and medial malleolus uniting.

Lateral view showing components well aligned and subtalar joint free.