Managing Non-Union of a Periprosthetic Distal Femur Fracture

    A decade after successful total knee arthroplasty, a 77-year-old female patient sustains a closed, comminuted supracondylar periprosthetic distal femur fracture in a slip and fall. Following failure of the initial treatment, a revision procedure is needed to achieve bony union.


    Brandon J. Yuan, MD


    A 77-year-old female patient presents following a slip and fall on the ice. She has history of non-insulin-dependent (type 2) diabetes mellitus, for which she takes metformin, as well as peripheral neuropathy and hypothyroidism. She underwent left primary total knee arthroplasty (TKA) 10 years prior. The knee had been functioning well before the fall.

    The patient sustained a closed, comminuted supracondylar periprosthetic distal femur fracture in the fall. Radiographs (Figure 1) demonstrate the comminuted metaphyseal distal femur fracture and what appear to be well-fixed femoral and tibial components.

    Figure 1. Radiographs taken after a slip and fall on the ice show a comminuted metaphyseal distal femur fracture.

    Periprosthetic distal femur fractures present several treatment challenges, including:

    • Lack of good quality bone in the short distal segment
    • Supracondylar comminution preventing direct reads for reduction
    • Femoral component potentially blocking placement of fixation implants

    Several studies have demonstrated suboptimal results when treating these fractures. This is related to the inability to obtain an appropriate reduction, implant placement, and high rates of non-union.

    Initial Treatment

    The treating surgeon chose to manage this fracture with open reduction and placement of a laterally based locked condylar plate. The reduction appears adequate on the postoperative radiographs: Although the plate placement may be slightly posterior on the distal segment, it otherwise appears acceptable (Figure 2).

    Figure 2. Postoperative radiographs demonstrate adequate reduction following open reduction and placement of a laterally based locked condylar plate.

    The presence of significant metaphyseal comminution prevents a direct reduction and, thus, the surgeon would be wise to opt for an indirect, functional reduction allowing for maximal preservation of the soft tissue envelope. An appropriate fixation strategy would be use of a construct that provides relative stability with some micromotion to allow for callus formation and secondary bone healing.

    In analyzing the postoperative radiographs, it is clear that the surgeon was conscious of the relative rigidity of the fixation. The working length of the plate has been increased, with the first screw in the proximal fragment above the fracture placed relatively proximal to the fracture itself. This is likely an attempt to allow for increased micromotion at the fracture site. Comminuted distal femur fractures stabilized with laterally based locked implants can often be made “too stiff” to allow for adequate callus formation due to the working length of the plate being decreased too much.

    On closer examination, a line of surgical skin staples traveling up the lateral aspect of the thigh – from the knee all the way to the proximal aspect of the plate – can be visualized (Figure 3). The presence of the skin incision does not necessarily mean that the underlying fracture has been exposed or surgically devascularized; however, it appears that an attempt at direct visualization and reduction of the fracture was made.

    Figure 3. Surgical skin staples traveling all the way up the lateral aspect of the thigh can be visualized on postoperative radiographs.

    The appropriate fixation strategy – relative stability – was selected, but it must be mated with the correct reduction and exposure strategy. The surgeon must be careful to avoid stripping the soft tissue at the fracture site during the course of exposure, which would harm the fracture’s ability to unite through secondary bone healing.

    The patient was made touch weight-bearing for the first 6 weeks postoperatively and then was allowed to bear weight as tolerated. Radiographs taken at the 6-week postoperative visit demonstrate interval callus formation (Figure 4). The fracture is proving to be biologically active, despite any stripping that may have occurred during the index operation.

    Figure 4. Interval callus formation can be seen on 6-week postoperative radiographs.

    However, at 6 months postoperatively, the patient reported increasing pain at the fracture site. Radiographs demonstrate mechanical failure of the implant at the fracture site without apparent loss of fixation proximally or distally (Figure 5). The patient was referred for revision surgery.

    Figure 5. Mechanical failure of the implant at the fracture site can be seen on 6-month postoperative radiographs.

    What Went Wrong?

    With any non-union, the surgeon should be concerned about infection, metabolic abnormality, and stripping of the fracture site as potential causes for the delayed healing leading to failure of the construct. In this case, the reduction and fixation construct appear to be appropriate for the fracture. In addition, the follow-up radiographs demonstrate interval callus formation, suggesting the fracture is metabolically active – but perhaps not metabolically active enough to have united prior to failure of the locked plate.

    Laboratory studies were obtained before revision surgery to assist in ruling out infection or metabolic abnormality. This included C-reactive protein, erythrocyte sedimentation rate, white blood cell count, and a panel of endocrine labs (vitamin D, thyroid stimulating hormone, parathyroid hormone, alkaline phosphatase, calcium, and phosphorus). All were within normal limits.

    The primary goal of this revision procedure was to create a mechanically and biologically favorable environment for union by addressing all potential shortfalls of the original operation. Given the concern for lack of robust vascularity and biologic activity at the fracture site, the plan was to add autograft bone to the non-union site. Options for the donor site include the iliac crest, the proximal tibia, and the femoral shaft, with the graft harvested via the reamer/irrigator/aspirator (RIA). The open box of the femoral component on this cruciate substituting TKA allowed for passage of reamers up to 15 mm. Thus, the intramedullary bone graft was harvested from the femur via a retrograde entry portal.

    Options for revision fixation include:

    • Revision lateral plating
    • Dual (medial and lateral) plating
    • Retrograde intramedullary nailing
    • A combination of techniques

    To fortify the non-union mechanically and preserve the medial soft tissue envelope, the decision was made to proceed with revision lateral plating combined with retrograde intramedullary nailing. Although the original fracture was significantly comminuted, the primary non-union site is often a single site or fracture line by the time of a revision procedure. These “more-simple” fracture lines require compression and increased stability to allow for union. To accomplish these goals, I choose the combination of techniques described (retrograde intramedullary nail combined with a laterally based locked plate).

    Revision Procedure

    • The patient was taken to the operation room and positioned supine.
    • The non-union was exposed and the previous broken plate was removed.
    • The fibrous tissue at the non-union site was debrided back to bleeding bone and samples were sent for culture (all negative for infection).
    • The distal femur was first stabilized with a new laterally based locked plate, with care taken to position 2 distal screws out of the pathway of the planned intramedullary device. In addition, 1 unicortical proximal screw was placed in the plate to provide provisional stabilization (Figure 6).

    Figure 6. A series of fluoroscopic images show the new construct: A laterally based locked plate secured with 2 distal screws, as well 1 unicortical proximal screw placed in the plate to provide provisional stabilization.

    • Intramedullary bone graft was harvested through use of the RIA and then a retrograde nail was placed and locked distally with 3 interlocking bolts. The remainder of the distal locking screws were placed into the plate, surrounding and interdigitating with the distal aspect of the nail.
    • Next, the single unicortical screw in the proximal aspect of the plate was removed, and the non-union site was compressed via impaction of the driving end of the retrograde nail (Figure 7).

    Figure 7. Fluoroscopic images of the non-union site before and after impaction of the driving end of the retrograde nail.

    • Following compression, the bone graft was added to the non-union site, the plate was secured to the bone proximally, and interlocking bolts were added to the proximal end of the nail (Figure 8).

    Figure 8. Final postoperative radiographs from the day of surgery. A broken drill bit is visible proximal to the nonunion site, from attempting to place screws around the intramedullary nail.

    Postoperatively, range of motion of the left knee was started and the patient was allowed to bear weight as tolerated with the use of an assistive device.

    Radiographs 6 months following the revision procedure demonstrate bridging callus and progression toward completely bony union (Figure 9). The patient has progressed back to baseline activity level without pain.

    Figure 9. Bridging callus and progression toward completely bony union can be seen on 6-month postoperative radiographs.

    Key Points

    • Comminuted metadiaphyseal distal femur fractures may be successfully managed with a retrograde medullary nail or lateral locked plate. However, basic principles should be obeyed: Comminuted non-articular fractures should undergo a functional, indirect reduction and should be stabilized with a construct that allows for micromotion through the zone of the fracture. With this approach, direct visualization of the fracture site is unnecessary, reducing the potential for further devascularization of the fracture site. Simple fracture or non-union patterns require compression and absolute stability if primary bone healing is to be expected.
    • Adding biologic supplementation to selected non-unions is an important consideration. Options include use of iliac crest, proximal tibia, or calcaneus bone. Bone grafts may also be harvested from the medullary canal of long bones with the use of the RIA.
    • In select cases, the use of a combined retrograde intramedullary nail and laterally based locked plate can assist in improving construct rigidity and anticipated load to failure in the distal femur.

    Author Information

    Brendan J. Yuan, MD, is an Assistant Professor in the Division of Orthopedic Trauma, Mayo Clinic, Rochester, Minnesota.

    Disclosures: Dr. Yuan has no disclosures relevant to this article.