Distal femoral fractures are complex injuries which do not tend to heal quickly. For closed injuries I tend to tell my patients that it will take at least 12 - 18 months for the fracture to heal. Open fractures with bone defects take even longer to heal and require special limb reconstruction techniques. Preventing complications, predominantly infection is of a paramount importance.
After the initial debridement of the wound and removal of obvious avascular bony fragments at the local hospital, the wound was closed (I prefer to close the wound if possible and then reopen it if necessary) and an external fixator applied. The patient was then transferred to our hospital.
The argument to leave all those comminuted fragments, or at least the bigger ones, in is reasonable however in my personal opinion it is much much easier to grow 1 cm of a new bone than deal(ing) with a non-union or even worse, infection. As you will see below and on the X-rays to come we are extremely aggressive in debridement of the bone and tend to take out all dead but also potentially dead and contaminated bone. Quite often we will find foreign bodies which we otherwise would not.
I always debride spikes on all fragments. One reason is that I do not know whether they were in contact with outside world. The next (other reason) is that the spikes will prevent sufficient docking during the callus distraction process. Again, it is much easier to grow 1 cm of bone then to deal with non-union or infection. The third reason is that they very likely represent (a) dead bone.
After the debridement and thorough washout are completed I will proceed with the reduction of the joint surfaces. Ideally I will use reduction clamps of any description and temporarily stabilise the fragments with K-wires. If possible I try and exhibit 0% tolerance to malreduction of the joint surface but not at the expense of the viability of the fragments or significantly lengthening the time of the procedure. You should know when to stop.
When joint surface is reduced two bone spike retractors (surgical instruments) are put on the proximal femur, one on the front and one on theback of the femur. When the plate (Stryker AxSOS) is inserted sub muscularly it should pass between the bone spikes. This confirms that the plate is in the position I want it to be. Exact confirmation of the plate position is not important at this moment in time.
The next step is to measure the length of the plate required. If possible I tend to measure it on the contralateral side, from the distal femur all the way to the tip of greater trochanter. I will even try and bend the plate to accommodate the greater trochanter but not all plates will allow bending.
The length is decided when the plate is inserted sub-muscularly all the way to the greater trochanter and confirmed with an X-ray. When X-ray of the distal femur with the plate is completed, the leg is pulled to restore the leg length. The correct plate length is confirmed.
At this stage the plate will have moved proximally so the next step is adjusting the plate against the distal femur. In the AP plane varus/valgus malalignment has to be adjusted. Even more important is to neutralise the flexion of the distal femur in the lateral view. When the distal femur is put exactly where it has to be against the plate, fixation of the distal fragment to the plate with locking and non locking screws is performed. It is very unlikely that limb alignment will be wrong at the end of the fixation if the distal femur has been positioned accurately against the plate.
On the X-ray images above you can see progressive correction of the flexion of the distal fragment until the fragment is perfectly aligned with the plate in the AP and lateral views.
Before we fix the plate to the proximal fragment we have to check the rotation of the leg. Currently we do this by comparing the projection of the lesser trochanters. The non-injured leg is put in the position where the foot is perpendicular to the table and an AP view of the (ipsilateral)lesser trochanter is taken and saved. Half pins inserted (from the damage control Ex-fix) into injured proximal femur are used to control the rotation of the proximal fragment until the projection of the lesser trochanter on the injured site is equal to the contralateral side. The injured leg is then positioned with the foot perpendicular to the table. This is then the position where the plate should be fixed to the proximal femoral fragment. I find this technique reliable and will only miss few degrees of rotational malignment. I personally tend to err toward excessive external rotation.
I work towards and accept shortening in the region of a half of an inch. There are few reasons to keep the leg a bit shorter. The first non-locking screw is inserted in a slightlyoblique direction (to add a bit to the length).
The position of the proximal fragment alongside the plate is confirmed and 3+1(the initial non-locking screw is replaced by a locking screw) locking screws are inserted in the proximal fragment, as proximally as possible.
At this point only definitive debridement of the medial condyle was left (I was not able to debride it at the beginning due to the difficult access but also mobility of the fragment). All my bone debridements are done using drill and osteotome (low energy break), I never use a saw to break a bone (high energy break).
When fixation and debridement is completed, the bone defect is packed with (a) bone cement in at least two-three layers (to increase cement surface area and enable easier removal). Anitbiotic beads are inserted between the layers of (the) cement and in the wound.
All wounds are closed in layers.
21 May 2016
Good progress in wound healing and mobility. No major concerns at the moment. X-rays show good position and alignment of fragments but also a bend on the plate suggesting the load the plate is under (I did not bend the plate).
20 June 2016
All wounds have healed. Half pin wounds on the proximal thigh not completely dry but certainly do not look infected. Decided to proceed with the second stage.
Prior to the second stage I had multiple discussions with Clinical Engineers at Addenbrooke's and Stryker Engineers from Switzerland.
Plan was to use a cable to move the fragment down and fill the gap (to minimise the the pin sites infections). From my previous experiences I know that two cables are strong enough but one is not enough to pull the fragment down. Using system of pulleys I can increase the power but slow down the speed of travelling. There are many other restrictions in the process but will not discuss them at this stage with the aim to keep it simple.
Intraoperative images are below.
Patient was discharged next day and will remain non-weightbearing. Surprisingly the pain was not bad at all. I have no explanation.
Postoperative X-rays are below.
05 July 2016
Pain is getting under control. Wounds are healing well. Clips removed. Has been transporting the fragment for the last week using half pins. All going well as you can see on the X-ray below. In two weeks time will remove the two half pins from the transported fragment and leave only the cable to do the job.
18 July 2016
Pain has improved. Also the knee joint movement has improved and is improving further. Two pins which were used for distraction of the fragment have just started to cause problems (skin irritation).
X-rays below confirm good regenerate and around 2 cm of distraction. No other untoward signs on the X-rays or clinical examination. Cable is tight and following the distraction. Distal femoral condyles have become atrophic which suggested that the blood supply is good and they are alive. Good sign.
In my opinion it is safe to remove the two distracting pins and let the cable do the job.
Two distal half pins were removed under sedation. The frame was dismantled and only one proximal clamp was left with the aim to keep the three proximal half pins as a one unit and to secure a point for the clicker. Wire was also repositioned in the threaded bar's slot.
X-rays below were done on the same day, just after the amendments were done. Cable is tight and hopefully two pulley system will be able to pull the fragment down with as little of distraction for the patient as possible.
26 July 2016
Continuous improvement. Pin sites clean, "cable" site as well. Still non weight bearing. X-ray below confirms that the cable moved the fragment distally for around 6-7mm. As expected. Maybe some reaction around the distal screw which acts like a pulley, but still looks stable and solid.
We also amended the "pulling" system as the cable was impinging on the skin.
Amendments were acceptable but the patient made much better and safer construct when at home. Really impressive.
23 August 2016
Slight delay in follow up but still satisfactory X-rays. Right knee range of movement is at the moment minimal. Will work on getting CPM machine and continue with physio. Hard work no doubt. Wire site clean.
As you can see from the X-rays below the double pulley system is working. There is slightly less of progress then would be expected after a month but this is mainly due to the need for changing position of the cable on the threaded bar. Around 5mm less then planned. Great so far.
02 September 2016
Problem with clicker as it was much tighter to turn then before. Otherwise knee movement is gradually improving. All three pin sites are clean and cable site clean as well.
Clicker removed but no fault found. It seems that the force required for the callus distraction or osteogenesis is significant.
Looking at the X-rays below and comparing with the X-rays from 10 days there is around 10mm more of distraction what is corresponding with the plan.
20 September 2016
Clicker remained the problem until it was replaced by the patient. Another issue was connection between cable and threaded rod. Again sorted by patient. After issues have been remedied the transport continued and as you can see it looks well on the X-rays below. In my opinion all the recent events confirm that patients should be fully engaged in the treatment.
Because of my worry of premature consolidation we are now moving the fragment 1.25mm per day. So far we have reached around 6 cm, only 10 cm to go. But to be honest, the defect did represent 1/3 of a femur. This does not sound massive, but trust me, it is.
Passive range of movement of the knee is satisfactory and is improving. Certainly no sign of infection and the cable site is clean. Will try and get a picture of it.
04 October 2016
Good progress. No changes. Pin and cable sites clean. Range of movement in the knee passively to 90 degrees, active very painful therefore not exercised. Since the last appointment we slowed down to 1mm per day. Regenerate is another cm longer, another 9cm to go.
01 November 2016
Going well. Proximal pin is still bending and causing some minor issues comparing to standard bone transport using mono lateral frame. Plate slightly bent comparing to the beginning but does not look like it is deteriorating. And probably less bent than off the previous X-ray. Rotation?
22 November 2016
Few days ago suddenly pain started at the back of the knee. X-ray shows that the new bone (calcified periosteum?) were pushed by the transported fragment behind the knee.
CT angio scan confirmed the position of the ossification, which is pretty central in the back of the knee and not to far away from the vessel. Not sure what to do next.
CT Angio coronal reconstructions
CT Angio coronal reconstructions
07 December 2016
Pain has improved and allowed further 12mm of transport but then came back. X-rays confirm further impingment at the back of the knee. After few days of rest pain has improved as well.
Otherwise no sign of infection, pin sites and cable site look clean. Certainly much better then what expected.
20 December 2016
Pain became back even stronger. Decided to remove the fragment from the back of the knee. Discussed different approaches. Finally decided for medial subvastus.
X-rays before the surgery (bellow) show further distalisation of the transported fragment but also the calcification.
Initial plan was to lock the transported fragment and accept the distal femoral non-union. During the surgery it became obvious that it would be relatively easy to shift the fragment for another 1.5cm distally and mimimise the gap and the need for docking site bone graft, hence the decision to conclude the surgery with removal of the calcification only.
Post operatively patient felt that the patinet has improved and moveing the fragment became much easier. Very pleased.
03 January 2017
Strength of the leg has improved dramatically. Unfortunatelyy the knee is still more or less unchangess. No problem with pin or cable sites. All wound healed.
10 January 2017
Yesterday all transporting metalwork (screw, cable and half pins) were removed, including the crimp. It was easier then expected.
Transported fragement was elevated using a clamp and locked in the position with 3 screws. Docking site was prepared and bone grafted using cancellous bone from the ipsilateral proximal tibia. Antibitic beads were used to sterilise the area. Despite very clean area, microbiology specimen were taken as per routine protocol.
Whole femur X-rays suggest that the original bending of the plate was definetily caused by the cable as the plate is back to the shape as it was before the bone transport started. Quite impressive what a force a plate can sustain.
20 January 2017
Microbiology results came back as positive (from samples taken) and antibiotic treatment has started as per microbiology advice. X-rays below do not show any significant changes. This is positive.
Clinically the leg looks fine, certainly not infected. Antibiotic beads should be doing the job.
10 February 2017
Good response on antibiotic. CRP almost normal. No clinical signs of infection.
Regenerate continues to consolidate, when docking site particularly medial part of it, is getting filled by bone a bit quicker than anticipated.
Plate and all screws will get replaced by new ones in four weeks' time allowing us and the patient enough time to prepare.
I am still not completely sure how are we going to do it?
06 March 2017
Prior to the procedure we took another set of X-rays to confirm the bony anatomy. Otherwise everything else was OK. Still on antibiotics.
In my opinion there is a good healing of the docking site, particularly medially, but laterally as well. Predicted?
For the plate exchange we manufactured pegs on advice from Stryker engineers. Pegs have proven to be very useful. One half pin per fragment external fixation was assembled. After we removed all of the screws from the plate we replaced screws with pegs and removed the plate. Prior to the removal of the plate we marked the holes which had the pegs inside. After the plate was replaced by a new one, the new plate was positioned according to the pegs and marked holes. There was a slight issue with the pegs lengths but were able to overcome it. At the end we got the new plate exactly on the place it was before and fixed it with the same size and length screws, apart from one which was 2mm longer. All screws were driven out and in on the bone threads suggesting no bone loss due to the possible instability. No instability noted on any of the X-rays either. New microbiology samples were sent.
14 March 2017
Wound was oozing few days after the latest discharge. As patient still on antibiotics and a lot of local antibiotics used decided to heal the wound wound with incisional VAC but keep a close eye on the wound.
04 April 2017
Wounds healed. Patient walking with crutches. Still under Infectious diseases consultant for antibiotics.
X-rays show good progress toward union in the docking site and consolidation of the regenerate.
Plate is bend. Unfortunately there is nothing we can do about it. The forces on the plate are quite impressive.
02 May 2017
Walking with crutches partial weight bearing.
Wounds healed, no regular painkillers. ROM 0-95 deg.
X-ray - further consolidation of the regenerate. Visible line at the docking site suggesting possible delayed union, when lateral view suggests filling of the gap.
16 June 2017
Walking weight bearing as tolerated. Wound healed, No clinical sign of infection.
X-rays suggested further healing in the docking site. Unfortunately still not complete union. Regenerate is continuously improving as well.
01 August 2017
There is a block on the knee movements at around 90 deg. Very likely due to the significant scarring of the vastus intermedium muscle. At this stage I am a bit worried about massive releases as it can damage the regenerate relatively easy. Will wait and see with the release.
CT scan does not confirm union. There is still a fracture gap visible. Whilst the docking site is not healing as expected, we have decided to wait for a bit longer as there is no signs of instability of the metalwork.
08 September 2017
Patient is walking unaided on shorter distances and with a crutch for longer distances. Still limping. Range of movement 0-100 deg of flexion. Reasonable but because of the nature of his activities this is not enough. Currently we cannot afford open release of quadriceps due to the regenerate, which is not completely ossified. There is also further progress on the non-union site. From my opinion the non-union is healing and there are some bony bridges visible on the X-ray. At the moment CT scan is not an option.
29 September 2017
No comments required. Disappointment.
05 October 2017
The broken plate was removed and exchanged with a retrograde femoral nail. To additionally stabilise the distal femur poller screws were used. As standard samples for microbiology were taken and Stimulan put in at the end of the procedure.
05 October 2017
Patient is full weight bearing, walking without crutches on shorter distances. Right knee range of movement is from 0-115 degrees of flexion. Certainly impressive. Result of a hard work and dedication.
Wounds healed and so far no sign of infection.
X-rays show further consolidation of the regenerate and very likely union of the docking site, particularly on the lateral side. Although I said that before and I was wrong. Certainly it does look promising. I will remain very positive.
19 December 2017
All wounds are healed. ROM from 0-90 degrees of flexion despite aggressive physio.
Patient is certainly not happy with it. In my opinion quadriceps release would very likely do more harm than not at this stage at least.
My suggestion was to continue with physio, maybe not as much on aggressive side than on a repetitive side.
X-rays don't show any deterioration of loosening.
15 January 2018
Just after Christmas patient felt sudden pain in the knee. X-ray revealed broken distal locking screw and slight subsidence of the distal fragment.
Patient was taken to theatre, broken screw removed and replaced. Additional two distal locking screws were inserted. The knee pain improved dramatically after the procedure.
06 February 2018
No significant pain or discomfort but the level of physiotherapy has been lowered to protect the fixation. The most distal locking screw is a bit too long and is causing some problem. We accepted it during the surgery because it had a really good purchase and I was worried in case of replacing it to loose the fixation.
Mr Matija Krkovic, MD, PhD
I am Consultant Orthopaedic Trauma Surgeon with special interest in Limb reconstructions and bone infections.