CORR Insights®: What Factors Are Associated With Implant Breakage and Revision After Intramedullary Nailing for Femoral Metastases?

Where Are We Now?

The surgical management of bone metastasis depends on many factors including the patient’s general condition, the histology of the tumor, the site and extent of the lesion, and the expected response to other treatment modalities such as radiotherapy and chemotherapy. Our goal is to palliate the pain and symptoms of bone metastasis, and to stabilize the bone for immediate weight bearing so that the patient’s quality of life is not unnecessarily impaired [8]. Unfortunately, bone healing and union in this patient population are not always guaranteed; local control may not be effective, there might be local disease progression [2, 3, 8], or the implants (particularly internal fixation devices) may break.

For those reasons, when we treat patients who are expected to live for more than a few months, we should use stabilization constructs or arthroplasty-based approaches that can survive longer than the patient. By contrast, metastases in patients who are not expected to survive for long should be managed in ways that minimize short-term complications. Intramedullary nails seem to meet the latter need well, in that they often can be inserted percutaneously in a relatively straightforward surgical procedure [3]. However, Weiss and colleagues [9] reported 11 reoperations out of 108 patients with proximal femoral metastases using intramedullary nails. It would be advantageous to know in advance which patients are at greater risk for that complication, since reoperation is something we especially would like to avoid in patients whose expected survival is measured in months.

In the current study, Willeumier and colleagues [10] helped us to answer that question. In a series of 228 femoral metastases managed with intramedullary nailing, the authors found that implant fracture was independently associated with nailing performed in patients with actual (rather than impending) fractures, and with prior radiation to the site. By contrast, patients who had received bone cement as part of the fixation construct were less likely to undergo revision surgery. Unsurprisingly, two thirds of the patients had died before 1 year, and more than 80% had died within 2 years of the procedure.

Where Do We Need To Go?

Could these findings guide us in our management plan? We most likely need to be more active in screening our patients with tumors for any risk of impending fractures according to prognosticating methods such as Mirels score [5] so that timely prophylactic fixation can be considered. In addition, we should avoid preoperative radiotherapy. Intraoperative use of cement minimizes the risk of revision as was shown here as well as by others [8]. However, we need to realize that the presence of cement may block the osteosynthesis permanently and its long-term effect for long-term survivors is unclear. For high-risk patients looking for better implant durability, we may consider alternative means such as an endoprosthesis, which may in fact be a megaprosthesis, depending upon where the lesion is located. However, the risk of infection and other complications is high in patients treated with prosthetic reconstruction [9]. Although failure of fixation is of course seen more in the intramedullary nail group, infection seems more prevalent in endoprosthesis patients; the timing of these events also is important when trying to choose between them. Deep infection tends to occur early and implant failures occurs later (longer than 6 months) [2, 3].

Additional gaps in our knowledge remain. For example, we still need to determine the likelihood of the bone healing with internal fixation when local treatment (radiotherapy) and/or systemic treatments are used. We also still need to improve the accuracy of our oncological survivorship predictions. In addition, Lipton and colleagues [3] recently found that bisphosphonates and more-potent denosumab lower the risk of skeletal related events, complications related to bone metastases including pain, pathological fracture, spinal cord compression, necessity for radiotherapy, and surgery to bone. Our surgical indications should continue to evolve in light of all of this new information.

Future studies might focus more on other patient-related risk factors. These patients may be heterogeneous in general health status, bone-healing potential, response to local radiotherapy, availability of systemic drug treatment, and life expectancy. Nonunion, local tumor progression, and the lifespan of a patient will likely determine whether an intramedullary nail is sufficient treatment for a metastatic lesion.

How Do We Get There?

The leader or coordinator of a potential multicenter study on the risk factors associated with implant breakage and revision after intramedullary nailing for femoral metastases could invite those centers that routinely offer long-term followup and have a computer patient care and retrieval system to be a part of the study. Other than patient consent for undergoing the study and detailed preoperative assessment, all other patient management including the indication, choice of implant, and intra and postoperative care should be the same as those not included in the study. The end points, such as implant failure resulting in further operation, and death, are major events that are usually captured in the hospital computer system and can be quickly retrieved.

Alternatively, a prospective randomized control study may be a viable way to analyze risk factors like preoperative radiotherapy and the use of cement. This type of study may be the only reliable way for a 10-year, single-center study to recruit enough patients, and we must anticipate that the oncological management of a patient could change within this time-period.

Filling up the defect in the femur bone with cement decreases relative motion between the two bone ends and ensures the nail as a load-sharing device. However, such a foreign body within the bone defect makes the bone union unlikely to occur. Theoretically, using cement reconstruction could result in the intramedullary nail breaking down. Still, Willeumier and colleagues found that more than 80% of these patients die within 24 months, a much shorter lifespan than an intramedullary nail and cement reconstruction. But in my opinion, patients will likely survive longer in future due to our medical advancements and indeed, the beneficial effect of using cement in augmentation may reverse in long-term survivors. A study including long-term survivors will help us attain more answers.

Although endoprosthetic replacement has less chance of implant failure, it is a more technically demanding procedure with a higher risk of infection and dislocation [9]. Thus, it discourages surgeons from agreeing on its routine use to manage the femur metastases. However, with the recent powerful drug treatment such as targeted therapy and new chemotherapy, long-term survivors may have uncomplicated bone union. In the future, pathological fracture may not differ from the traumatic fracture of the femur bone. In that time, intramedullary nails without cement augmentation may be the first choice in surgery selection.

There are studies evaluating the prediction of oncologic survival among patients with spinal bone metastasis [1, 6, 7]. The Bollen prediction model for survival [1] is a useful tool for orthopaedic surgeons in managing spinal bone metastasis. The data in this model includes primary cancer, performance status, presence of visceral, brain and bone metastases, number and locations, and neurological status. A study predicting patient survival for the femur bone metastases may be warranted. However, models similar to the Bollen prediction model are only a reflection of patients' survival during the study period. In this era of rapid advancement in medical science, the prediction may not be valid after a few years. It seems that regularly repeating the study is the only way to give a reliable answer to the patients' survival.