Fixation Failure in an Atypical Femur Fracture

    A 59-year-old female patient presents with a stress fracture at the lateral cortex of the proximal femur. Initial fixation with reconstruction nailing fails, as does revision with a blade plate and autologous bone grafting. Will a second revision procedure allow for union of the fracture?


    Brandon J. Yuan, MD

    Case Presentation

    A 59-year-old female patient with a history of rheumatoid arthritis, osteoporosis, and pulmonary embolus presents with activity-related right thigh pain of 3 months’ duration. She had been treated with bisphosphonates for osteotoporosis in the past but has not been on medication for bone health, other than vitamin D and calcium supplementation, for more than 5 years. In addition, she had taken prednisone for rheumatoid arthritis in the past.

    Radiographs were obtained and demonstrated a stress fracture at the lateral cortex of the proximal femur (Figure 1). Note the periosteal hypertrophy and beaking, as well as the “black line” consistent with a stress fracture and concerning for impending completion of this atypical fracture.

    Figure 1. Stress fracture at the lateral cortex of the proximal femur showing periosteal hypertrophy, beaking, and a black line.

    Major and minor features of atypical femur fractures have previously been described:

    • Major features present in this case include the incomplete fracture on the lateral cortex without a history of trauma.
    • Minor features present include localized periosteal reaction, prodromal thigh pain, and comorbid conditions (rheumatoid arthritis in this case).
    • The use of bisphosphonates and glucocorticoids have also been associated with an increased rate of atypical femur fractures. [1]

    This constellation of radiographic signs and symptoms of activity-related thigh pain may warrant consideration for prophylactic fixation of the femur. Certainly, a stress fracture visible on plain radiographs through the lateral cortex increases the risk of completion of the fracture if prophylactic fixation is not performed.

    The physician subsequently ordered an MRI of the femur despite the fracture being visible on plain radiographs. In the intervening time, the fracture completed with minimal trauma.

    Initial Procedure

    The patient was taken to the operating room for closed reduction and reconstruction nailing of the right femur. Unfortunately, intraoperative fluoroscopic images are not available.

    The patient continued to have significant thigh and groin pain postoperatively. At the 1-month visit, the patient was given instructions to continue to ambulate with crutches despite the radiographs obtained at the visit demonstrating complete failure of the fixation (Figure 2). Failure was primarily due to the fact that the interlocking reconstruction screws do not pass through the nail. A true cross-table lateral radiograph was not obtained at that time.


    Figure 2. Complete failure of fixation after the initial closed reduction and reconstruction nailing of the right femur.

    The patient presented to my clinic 6 months postoperatively. Radiographs performed at that time confirmed that the proximal interlocking bolts did not pass through the reconstruction nail (Figure 3).

    Figure 3. Proximal interlocking bolts did not pass through the reconstruction nail.

    What Went Wrong with the Initial Procedure?

    The use of an antegrade medullary reconstruction nail is an excellent option for treatment of subtrochanteric femur fractures. All modern intramedullary nail systems have an aiming arm to guide proper placement of the reconstruction screws into the femoral head. However, the surgeon should always be aware that the use of an aiming arm is not fool-proof.

    Utilizing the aiming arm to forcefully rotate the nail, especially after it is already been seated fully, can introduce a small amount of error in the aiming arm system, causing the drill to miss the nail. Seating the nail by striking the aiming arm without the appropriate adapter or attachment can cause this error to occur. In addition, the surgeon or operating room staff should always check that the aiming arm truly does guide a drill bit perfectly into the interlocking holes in the nail prior to introducing the nail into the patient.

    At this point, the patient has a non-union of an atypical proximal femoral fracture with varus deformity and bone loss in the proximal fragment due to the presence of the medullary nail, with a nearly completely unstabilized fracture. Options for managing the patient at this point include the following:

    • Revision internal fixation with bone graft, utilizing a plate. Plate options include a proximal femoral locking plate or a blade plate. A blade plate has the advantage of allowing for excellent control of the proximal fragment to correct varus and regain length, in addition to providing the opportunity for direct compression of the lateral femoral cortex through tensioning of the plate.
    • Revision internal fixation with bone graft, utilizing a medullary nail. Revision nailing would also allow for correction of varus and length. It also has the potential benefit of the implant being positioned in line with the anatomic axis of the femur, a more mechanically advantageous position. However, this patient has significant bone loss around the trochanter and the inferomedial aspect of the proximal fragment, which could compromise the fixation of the nail in the short proximal fragment.

    Prior to any revision procedure, the surgeon must rule out underlying infection or endocrine abnormality that could contribute to delayed bone healing. Preoperatively, 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 – were within normal limits. This patient was also referred to an endocrinologist for consideration of use of anabolic agents to stimulate bone growth (eg, teriparatide), given her history of previous bisphosphonate use, rheumatoid arthritis and steroid use, and the atypical femur fracture. The patient was not started on any anabolic agent at this time.

    Revision Procedure

    Goals of the revision surgery were to:

    • Remove the prior intramedullary implant
    • Mobilize the fracture
    • Correct the varus deformity and shortening
    • Compress the fracture
    • Supplement with autogenetic bonegraft

    At the time of revision surgery, a drill path for the previous proximal interlocking bolts was noted anterior to the nail (Figure 4).

    Figure 4. Drill path for the previous proximal interlocking bolts anterior to the nail.

    Following removal of the prior implants, the fracture was mobilized and the proximal fragment was controlled with a blade plate (Figure 5).


    Figure 5. The proximal fragment was controlled with a blade plate.

    My initial thought was that use of a blade plate would allow for the most effective correction of the deformity and compression of the lateral cortex, both of which are key elements in successful healing of a subtrochanteric fracture.

    The fracture site was compressed using the blade plate and the articulated tensioning device. The non-union was supplemented with autologous bone graft from the proximal tibia. Postoperative radiographs demonstrated restoration of appropriate proximal femoral valgus and limb length (Figure 6).

    Figure 6. Restoration of appropriate proximal femoral valgus and limb length postoperatively (top). Radiographs of the ipsilateral hip prior to fracture are shown for comparison (bottom).

    The patient initially did well postoperatively. However, she subsequently developed recurrent right thigh pain and presented back at 9 months postoperatively.

    Radiographs demonstrate failure of the initial construct and progressive varus malalignment (Figure 7). Although excellent alignment was maintained for several months, eventual mechanical failure occurred through fracture of the distal screws.

    Figure 7. Failure of the initial construct and progressive varus malalignment.

    What Went Wrong with the Revision Procedure?

    The bending forces on a fracture in the proximal femur are substantial. Clearly, the tension on the lateral side – which was part of the reason for the development of the fracture in the first place – was enough to fracture the distal screws despite anatomic initial alignment and tensioning of the plate.

    It is well known that atypical femoral fractures are more likely to go on to non-union and take significantly more time to unite than standard subtrochanteric femoral fractures. [2]

    Although the blade plate allows for excellent control of the proximal fragment for correction of deformity, it places the implant along the lateral aspect of the femur, away from the mechanical and anatomic axis of the bone. The implant is unequally loaded and at risk for failure if the bone does not heal.

    Given this, it may have been better to use a medullary implant to allow for more even load distribution between the patient’s bone and the implant, thus leading to a longer time to potential implant failure and, eventually, union.

    In any case of failed fixation of fracture fixation, a standard preoperative lab workup is obtained, including:

    • C-reactive protein
    • Erythrocyte sedimentation rate
    • White blood cell count
    • Endocrine labs
      • Vitamin D
      • Thyroid-stimulating hormone
      • Parathyroid hormone
      • Alkaline phosphatase
      • Calcium
      • Phosphorus

    All were unremarkable in this patient. However, the patient was started on teriparatide to help stimulate bone growth given the persistant non-union and her medical history.

    Second Revision Procedure and Outcome

    A second revision surgery was planned. To increase the durability of the construct, a medullary implant was placed (Figure 8). A large medullary nail was utilized in an attempt to fill up as much of the space as possible in the canal of the subtrochanteric femur to assist with alignment. The larger implant also allows for increased construct rigidity.

    Figure 8. A medullary implant with a large medullary nail was placed to increase the durability of the construct and assist with alignment.

    Due to the patient’s history of stress fracture and failed fixation, the biology of the fracture site was augmented as well. Iliac crest autograft bone was placed at the fracture site to help stimulate new bone formation.

    The patient’s weight-bearing was protected for 6 weeks, followed by transition to full weight-bearing. Her fracture went on to uneventful union, as seen in radiographs obtained 1 year after surgery (Figure 9).

    Figure 9. Radiographs from 1 year after the second revision procedure.

    Key Points

    • Atypical femoral fractures represent unique entities. Patients with these fractures often have coexisting medical or endocrinologic abnormalities that impair bone healing. In addition, it has been demonstrated that these fractures are less forgiving to malalignment in varus. They take significantly more time to heal and are more likely to go on to nonunion.
    • When using the aiming arm of the nail to guide placement of interlocking bolts, ensure that the aiming arm fits appropriately and truly does guide a drill through the nail prior to nail insertion. Avoid forceful twisting of the nail when using the aiming arm during nail insertion. If the nail must be seated by striking it with a mallet, make sure the appropriate attachment is utilized.
    • Given the knowledge of prolonged healing times with atypical femoral fractures, consider the use of a medullary implant when feasible as the implant durability may be superior to the use of a laterally based plate.
    • Adding biologic supplementation to selected non-unions is an important consideration. Options include use of iliac crest, proximal tibia, or calcaneus. Bone graft may also be harvested from the medullary canal of long bones with the use of the reamer/irrigator/aspirator.

    Author Information

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


    The author has no disclosures relevant to this article.


    1. Edwards BJ, Bunta AD, Lane J, et al. Bisphosphonates and nonhealing femoral fractures: analysis of the FDA Adverse Event Reporting System (FAERS) and international safety efforts: a systematic review from the Research on Adverse Drug Events And Reports (RADAR) project. J Bone Joint Surg Am. 2013 Feb 20;95(4):297-307.
    2. Bogdan Y, Tornetta P, Einhor TA et al. healing time and complications in operatively treated atypical femur fractures associated with bisphosphonate use: a multicenter retrospective cohort. J Orthop Trauma. 2016 Apr;30(4):177-81.