Sleeved MBT Total Knee Replacement for Tibial Plateau Fracture (DePuy)

This 64 year old female patient sustained a Schatzker II tibial plateau fracture, a lateral plateau fracture with central depression, in a low energy simple fall.
Schatzker II are the most frequent type of injury I personally have treated with the Depuy Sleeved MBT TKA.

This post casting radiograph clearly reveals that the knee is unstable with valgus deformity. However whilst in the ED it was documented that the medial structures were intact.
Patients should be managed in either a ROM brace or cast brace to allow passive and active flexion exercises. I would not permit flexion past 90 degrees (weak evidence base that flexion past this causes further impaction of the fracture)
This patient presented to me 2 weeks following the injury as a tertiary referral. In my opinion 2 weeks is satisfactory to allow the soft tissue envelope to heal. Delays past this result in stretching of the medial collateral ligament and fibrosis of the lateral structures.

This series of radiographs reveal the problem with ORIF in osteoporotic patients. The fracture is multifragmentary and the impaction of the lateral femoral condyle into the lateral tibial plateau has resulted in a significant cavity.
Restoration of the joint line, bone grafting and reduction and fixation of the multiple fragments would have at best been problomatic.

An pre-operative examination under anaesthesia is essential. Here, this revealed a significant valgus deformity which was not fully correctable (secondary to the delays in surgery). The medial collateral ligament was intact.
Therefore before ‘knife to skin’ it is apparent that a fully constrained knee (hinge) is unlikely to be necessary but that soft tissue releases are likely to be required.

The surgical approach is the same as for a standard total knee replacement. The skin incision is from approximately 2cm above the patella with the knee in 90 degrees of flexion to just medial and distal to the tibial tubercle.
Once through fat, it is essential to identify the sub-bursal plane. This can often be difficult in patients who have sustained recent trauma as haematoma and inflammation make identification challanging.
Dissection is deep to the bursa. Focus should be on the medial side and significant lateral dissection should be avoided.

As for all valgus knees, the medial sweep should be limited. It is not necessary to extend posteriorly around the medial plateau. Had this been done, soft tissue balancing in the presence of tight lateral structures would have been very difficult.

The fat pad directly under the patella ligament is released. Ensure the blade is oriented caudally. A simple slip if oriented cephalically places the ligament at risk.

I preserve the infrapatella fat pad in a total knee replacement. Whilst data remains limited, two recent meta-analyses have shown that patients who have had excision of the infrapatella fat pad are more likely to exhibit anterior knee pain.
Different Factors Conduct Anterior Knee Pain Following Primary Total Knee Arthroplasty: A Systematic Review and Meta-Analysis. G Duan et al. Journal of Arthroplasty 2018
Influence of the Infrapatellar Fat Pad Resection during Total Knee Arthroplasty: A Systematic Review and Meta-Analysis. C Ye et al PLOS one 2016

Following release of the ACL, the PCL is cleared using diathermy from the femoral notch. To check it has been released completely the forceps can be inserted to confirm.

In the trauma situation the rotation of the tibial tray may be problematic. For the Schatzker I-III fractures, the lateral border of the medial plateau can be marked out (see slide 13). A horizontal line drawn vertically on the anterior tibia then marks the centre of the knee. This usually corresponds to the the inner 1/3 & 2/3 junction of the tibial tubercle but not always. I prefer to rely on the diathermy marking over tibial tubercle marking when there is discrepancy.

For tibial plateau fractures the standard PFC instruments can be used. However the tibial cutting jig has a 2 degree posterior slope.
I find the half blocks more versatile as they can placed closer to the bone when the fat pad has been retained.

Obviously with a lateral plateau fracture there is no lateral referencing. I chose to take 6mm from the medial side.
Once resected the plateau is removed with sharp dissection. With a posterior sacrificing system there is no need to preserve the PCL insertion on the posterior aspect of the tibia

This shows the resected plateau. As can be seen, there is significant destruction of the lateral plateau. Much of the articular surface was impacted into the metaphysis.
Photography of this area proved too difficult as sterility was jeopodised.

The next step is to create space for the tibial stems. An opening 9mm drill is used and advanced to 4cm. This needs to be in the midline of the canal and ANTERIOR. The midline will have been marked with the diathermy (as per slide 10 & 13).
A common mistake is to posteriorise the entry point. This results in posterior overhang when the definitive implant is inserted. The MBT stemmed tray is not modular in the AP plane. Therefore the correct entry point is paramount.

Once the pilot drill has created a channel, it is expanded using hand reamers. Different length stems correspond to different marks on the reamer. For a 115mm length stem the reamer is advanced to halfway between the 3rd and 4th marking.
Another note of caution. As you will see from slides 39-45 the tibial implant is constructed in the following sequence; the tray and sleeve are impacted and the stem is then screwed home. However, if revision of the implant is required the tray and stem are removed first and the sleeve is then removed. A stem of greater than 14mm will not fit through the central hole of the sleeve. Removal is therefore extremely difficult and I would advise that if considering a stem diameter greater than 14mm I would instead choose a longer thinner stem.
As mentioned earlier, we do not work in an ideal world. In an ideal world the correct length and diameter stem would not usually be dropped on the floor.

The natural flow of the TKA was somewhat disrupted whilst another set of trial stems was located and opened. Therefore we turned our attention back to the femur.
The pilot hole for the femur is made (as for any PFC IM rod guide) just above the femoral notch and medial. I find drilling at the angle shown in the slide until the cartilage is breached and then aligning the drill prevents slipping.
I would also recommend that you palpate the distal femur when inserting the drill and pay particular attention to the amount of flexion and extension of the drilled hole. Look laterally to ensure correct angle.

The cutting jig is attached to the IM rod which is inserted into the femur. In a non-complex knee I use a 5 degrees valgus angle setting. This represents the difference between the anatomical axis of the femur and the mechanical axis of the femur (centre of femoral head to centre of intercondylar notch) The handle is turned vertically and the two pins are impacted into the distal femur.
As a tip here, the distal femoral cut will be at right angles to the long axis of the femur (in the sagital plane). However errors can occur resulting in either a flexed of extended position. If the cutting jig is placed in the correct rotational alignment (as referenced in the procedure “Vanguard Total Knee Replacement) errors of flexion and extension will not additionally result in errors in the valgus or varus plane.

The amount of bone resected for a simple primary knee us usually 10mm. This is acheived by squeezing the black button (A) on the distal femoral cutting block and sliding it along the femoral outrigger (B). An ‘angel wing’ inserted into the distal cutting slot and should just touch the intercondylar notch (as indicated by the forceps).
The height of the outrigger is adjusted by the anterior button (C)
If the distal femur is highly eburnated (significant cartilage loss) I reduce the amount of bony resection to 8mm. This will avoid raising the joint line.

The distal femoral cut is made and is made easier with the assistant pushing back on the tibia (this elevates the femur).
Please note the Hohmann retractor is placed between the MCL and the bone of the femur, protecting it from inadvertent damage from the oscillating saw.

There remains much debate concerning obtaining the correct orientation of the numerus axes for femoral component positioning. I would refer the reader to the article by Claudio Castelli et al who reviews this issue in detail. There are many variations of the correct axis. I personally use the intercondylar axis as my rotation guide.
Rotational alignment of the femoral component in total knee arthroplasty. C Castelli et al. Annals of translational medicine 2016

The femoral sizing guide is then used to measure the femur. It is placed flush against the distal femoral cut. It is essential that this distal cut is correct as sizing errors occur when the cut is in extension or flexion (undersize and oversize respectively).
A stylus extends along the femur lifting the styus arm. When the size is determined, lock the guide (A). I always reference from the lateral femur as this is higher and there is therefore less chance of notching.
There are two types of guide with the PFC kit.
The anterior sizing guide will ensure the the anterior flange of the prosthesis lays flush with the anterior cortex of the femur, this resects 8mm of the posterior lateral condyle.
The posterior referencing guide references from the posterior condyles and likewise still ensures 8mm is taken from the posterior condyles.
Both work well when the size of the implant is an exact fit however patients may be ‘between sizes’
If a femur is between sizes different approaches to optimising the flexion gaps are required
For anterior references downsizing results in excess removal of bone posteiorly, resulting in an unstalbe flexion gap.
With posterior referencing downsizing results potentially in notching.
Therefore to correct for this in anterior referenced knees, move the drill guide slightly posteriorly and the reverse for a posterior reference guide.
Once the size is set (from A) the central portion of the guide is moved to the corresponding size and tightened (B). Please note this photograph was taken before the sizes in A &B were set
Finally pins are impacted into the drill holes pins (C), which when removed, create holes for insertion of the relevant sized cutting block.
The pins are asymmetric and position the chamfer cutting block in 3 degrees of external rotation.
This results in resection of 8mm from the posterior lateral condyle and 10-11mm from the medial. Again as the arc of radius of the medial condyle is greater more medial bone required resection to generate a ‘rectangular flexion gap.’
Details of this are documented in Depuy’s op-tech (please see link in the ‘implants’ section).
A good review of alignment in total knee replacement can be found in the paper by Cherian et al.

Mechanical, anatomical and kinematic axis in TKA: Concepts and Practical Applications. Cherian et al. Curr Rev Musculoskelet Med. 2014 Jun; 7(2): 89–95

The chamfer cutting block is placed in the guide holes created from the sizing guide and impacted home. Ensure it is flush to the bone on both sides. Errors here lead to knee imbalance.
The arc of curvature of the medial femoral condyle is larger than the lateral and a corresponding greater amount of posterior medial femoral condyle is resected with the PFC knee system. This should be confirmed before resection.
In practice the medial drill hole will be slightly above the intercondylar line and the lateral slightly below.

Before inserting retaining pins into the chamfer block ensure that the cuts will not notch. Test both the medial and lateral side. As the lateral side is higher, notching is more likely to occur on this side.

I personally cut the medial side first. It the cut is too posterior the chamfer jig can be anteriorised (confirm correct internal / external orientation first). Cutting incorrectly laterally first leaves a larger bony defect.

The remaining chamfer cuts are made. Again use the Hohmann to protect the colleratel ligaments

As mentioned the natural flow of this operation was impeded however attention returned to the tibia.
When similar situations occur (and they do in revision surgery when plans change on table) I omit the box cut (see slide 32) as excessive leavering on the condyles with the bone from the box removed can lead to fracture.
The tibia is prepared to take the metaphyseal sleeve with the tapered reamer. These increase in size sequentially.

Once reamed, the metaphyseal bone is prepared with the MBT revision broach. These are asymmetrical, therefore ensure that the ‘anterior’ laser etching is seen.

The broach is impacted until it’s superior surface is level with the tibial cut

Further cuts can be made using the superior surface of the broach as a guide. The tibial saw is often too large in small knees and a ‘shoulder’ type saw is often more practicable.
Of note in this case, I did not oversize the broach. The tibial fracture had not fully united and “overstuffing” it would have resulted in fracture displacement / propagation.

The tibial trial consists of a stem, a trial sleeve and tray.

I have found that the easiest way to ensure correct orientation of the tibial is to insert the sleave trial into the tibia first.

The tibial tray and stem are then inserted through the sleeve. Orientation is achieved before impaction. the cauterized reference line is still visible.
Of note, the lateral defect void was still apparent after tibial resection. Examination of the fragments revealed that fibrous union had occured in places. The cavity confies were secure enough to allow some limited impaction bone grafting.
This has not be elaborated upon here as impaction bone grafting is a whole case in itself and will be covered later.

Finally (and somewhat out of sequence) the box cut for the femur was completed. The tray acted as a barrier to prevent tibial penetration, femoral elevators are available or simply place the ‘angel wing’ under the saw tip.

The femoral component was impacted home. To insert the femur I have the femur mounted on an impaction device as I think this gives more control than holding the femoral component simply in the hand.
The final punch is anteriorly to ensure no flexion defect.
The trial spacer is inserted (see slides 48 & 49 for details)

It can be seen from this slide that the knee is not balanced. There is laxity on the medial side and a tightness laterally. The most likely reason for this is displacement of the lateral tibial plateau, which tensioned the lateral structures.
However although not united the fracture fragments were not mobile. Dissecting these out could have resulted in avascular necrosis of the fragments, further bone loss or iatrogenic nerve damage.
The on table decision was therefore to perform soft tissue releases.
One could argue that when performing soft tissue releases that a more constrained implant should be used however a deeper box cut results in removal of more healthy bone. Keeping as much healthy bone stock is essential, especially if revision is required (I always think of the next operation as a revision surgeon!).
The box cut for the PFC and TC3 is constant across all knee sizes (not so with the Attune), therefore the ‘box to bone’ ratio in this small size two femur would have been large.

I would strongly recommend the reader interested in soft tissue balancing of TKAs to obtain a copy of “Ligament Balancing in Total Knee Arthroplasty” by Leo Whiteside, Springer, ISBN 978-3-642-63924-1.
Firstly the posterior capsule was released without any significant effect on the deformity.
The next structure to be released was the ITB. This can be achieved from inside the capsule however external extra-articular dissection has been widely reported (and is covered at length in the book above).
Again this did not achieve the full effect.
It was noted that the lateral structures were predominantly tight in extension. Therefore release of popliteus (tight in flexion) was not required.

Unfortunately it is difficult to see LCL release from this slide. I will however be releasing some cadaveric workshop images soon. Getting close to the releases with a camera does jeopardize patient safety.
However as the knee was predominantly tight in extension the posterior aspect of the LCL required release. This was achieved by starting superior to the epicondyle and dissecting in a anteroinferior direction.(A) (popliteus is a more anterior and inferior structure and is released by working from an anterior to posterior direction).
Rechecking with both the knee flexed and extended revealed that balance had been achieved. Therefore the need for a more constrained device was not necessary.
However the size of the spacer required to maintain adequate tension was large

One concern with ever increasing tibial spacers is that the joint line is raised.
On table assessment of joint line is important. It can be difficult in the revision setting but in a primary knee common landmarks used include referencing the joint line height from the resected meniscal remnant or assessing the joint line from the inferior pole of the patella. (one finger breadth is a good rule of thumb!).
Alternatives include referencing the joint line 3cm distal to the medial epicondyle and 2.5 from the lateral or 12-16mm distal to the femoral PCL origin.
The error with these however lies in correct identification of the bony landmarks
In this instance increasing the tibial spacer resulting in an elevation of the joint line. Why this occurred is most likely secondary to an asymmetric stretching of the MCL. The femoral cuts were rechecked and did not seem excessive.

To bring the femoral component down, two 4mm distal femoral augments were used. These screw into the directly into the distal femur with a torque screw driver.
There is some debate as to whether a femoral stem should be used in when distal augments are used. It is my personal feeling that whilst sheer stresses will be increased on the cement bone interface I am not convinced that a stem is needed.
Firstly the presence of the box (although made shallower by the presence of the augments) still provides a better ‘hold’ than the 2 lugs on a cruciate retaining implant.
Secondly, forces through the femur are lessened as the knee system is a mobile bearing with a rotational interface between tibial tray and spacer.

There is some evidence to support this.
Anderson J et al. Constrained condylar knee without stem extensions for difficult primary total knee arthroplasty. J Knee Surg 2007;20:195-8

This image shows the partially coated porous metal sleeve. With an intact contained tibial defect press fit is available, however this is not possible in the setting of a tibial plateau fracture.
Depuy have stated that 60% integrity is required to get a satisfactory grip. This should be estimated from axial CTs prior to surgery.
However I have had successful outcomes in situations when less than 60% is available. A revision hip and knee surgeon needs to be adaptable (but remain sensible).

There are two ways of gauging correct alignment of the tibial tray within the sleeve.
The first is to copy the alignment created from the trials.
The second and my preferred is to insert the definitive sleeve into the prepared tibia. Care is taken not to advance it too far (as it will never come out)

The definitive tray is then gently placed inside the sleeve. Correct orientation is then dialled in. Finally a stout ‘tap’ with the fingers secures the mortise joint sufficiently to remove the construct as one.
Note the sleeve / tray construct can accommodate 20 degrees of internal or external rotation

The tray and sleeve are inverted on a firm padded surface. The anaesthetist is warned that there is likely to be loud banging (important if patients have only sedation and spinal anaesthesia).
The sleeve adapter is placed over the sleeve and the construct impacted firmly.

The stem is screwed home and tightened using two wrenches. the lower grips the undersurface of the tibial tray and the upper is tightened over the flutes of the stem.

As soon as the cement is mixed I smear a thin amount over the implants. Care must be taken to avoid the porous coating of the sleeve.
There is some emerging evidence that early application of cement to implants before the cement has entered the working phase increases the chemical bonding of cement and implant.
Conclusive data is however not yet apparent.

Once the bone cement is at its working phase it is applied to the washed and dried bone. It is important to remove loose debris and fat. Sclerotic areas should be overdrilled.
I still apply the cement in a total knee replacement with a spatula however I pay particular note to force cement down hard into the bone. The use of a cement gun (which achieves the same effect) is becoming more popular. Cement should penetrate at least 3mm into the cancellous bone

Tthe femoral component is impacted into place and excess cement cleared.

The tibial spacer consists of two parts. Firstly a trial revision post is inserted into the tray.

The trial spacer slotted onto this.
The knee held in extension (and very still) until the cement is cured.

Patella tracking is important and I ensure that the patella is able to be moved perpendicular to the femoral trochlear grove without undue tension. If this is not possible releases may be required

Once cured the final definitive tibial insert is introduced. The insert has an post on its undersurface with slides into a corresponding hole in the tray. It is therefore able to rotate (the mobile bearing!)

The insert sits flush on the tray

I close the arthrotomy with 1 vicryl. In my opinion it is important not to take large ‘bites’ of the patella ligament. Large bites when tightened will cause patella ligament shortening. I also believe that the vicryl is an irritant which may lead to posterior ligament neovascularisation. This has been associated with anterior knee pain. As in many things in orthopaedics I do not have any evidence for this but to me it appears logical.

I close the fat in two layers (unless the patient is very thin) and close skin with subcuticular monocryl.
I use glue for the skin closure in these cases and have found it seems to reduce wound leakage.

Post op images of the knee.
Key finding on this image are that the acrylic cement has not penetrated into the lateral bone graft site. Impaction was acceptable.
There has been no fracture propagation
Orientation is acceptable (taken within the context that these are not true AP and lateral films).
The fluted stem has good cortical hold.
The joint line (when scaled and measured) is not elevated.

Clinical photographs showing the expected range of movement for this procedure.
The preoperative deformity has been fully corrected.
These images were taken at the 8 week follow up appointment.