Treatment of Metastatic Carcinoma and Myeloma of the Femur
Joint MSTS/ASTRO/ASCO Guideline
Imaging and Clinical Findings
In the absence of reliable evidence, it is the opinion of the workgroup that the combination of imaging findings and lesion-related pain is predictive of risk of pathologic femur fracture. There is no reliable evidence to suggest that MRI is a strong predictor of femur fracture.
Although advanced 3-dimensional imaging, multi-plane x-rays, and combinations of studies, including PET scanning (Ulaner, 2017) can demonstrate radiographic depictions of the damage caused by metastatic lesions to the proximal femur, the available literature fails to define specific parameters that can accurately predict fracture risk. Low-quality evidence (Oh, 2017; Ulaner, 2017) supports the intuitive presumption that increased bone damage in the proximal femur is associated with an increased fracture risk. Furthermore, although MRI evaluation can accurately demonstrate the intra and extraosseous extent of lesions, there is no reliable evidence that this modality can be used as a predictor for fracture. Combining clinical factors, particularly tumor pain and pain with weight bearing, may aid clinicians in deciding when to intervene surgically in order to prevent a frank pathological fracture and the associated morbidities which may then occur.
Efficacy of Bone Modifying Agents
In the absence of reliable evidence, it is the opinion of the workgroup that the use of BMAs may assist in reducing incidence of femur fractures in patients with bone lesions from metastatic carcinoma or multiple myeloma.
We did not identify any literature with the purpose of determining the efficacy of BMAs in reducing femur fractures or other skeletal related events specifically among patients with metastatic carcinoma or myeloma lesions involving the femur. However, there are studies among patients with metastatic carcinoma or myeloma bone lesions not explicitly localized to the femur (Fizazi, 2009; Hortobagyi, 2017; Raje 2016/18; Lipton 2000/12; Morgan 2011/13; Stopeck, 2010, Martin 2012) demonstrating reduction in skeletal related events with use of BMAs.
Due to the observed benefit in these studies of improved clinical outcomes in context of the acceptable safety profile of commonly used BMAs, it is our consensus that treatment with BMAs in patients with metastatic carcinoma or myeloma involving the femur is advised.
Future research should specifically assess outcomes of femur fractures in patients with metastatic carcinoma or myeloma treated with BMAs.
Doseage Response of BMAs
Clinicians should consider decreasing the frequency of zoledronic acid dosing to 12 weeks (compared to the standard 4-week interval), as this is associated with non-inferior SRE outcomes and similar adverse event rates in patients with metastatic carcinoma or multiple myeloma. Clinicians should consider long-term use of BMAs to reduce skeletal related events in patients with multiple myeloma.
The question of zoledronic acid (ZA) dosing interval, i.e. less frequent dosing, has been addressed in several non-inferiority trials, in breast cancer patients (Hortobagyi, 2017; Amadori, 2013), and in a heterogeneous cohort of patients with multiple myeloma and metastatic carcinomas (Himelstein, 2017). These studies compared ZA 4mg dosed every 4 weeks to every 12 weeks, either upfront or after 12-15 months of 4-week ZA (Amadori, 2013). In each study, SRE rates were similar between groups, as were adverse event rates. In one study including myeloma and breast cancer patients, ZA 4mg IV was found to be superior to pamidronate 90mg IV (Rosen, 2004). ZA also has established efficacy in patients with non-small cell lung cancer and solid tumors other than breast and prostate carcinoma (Rosen, 2004). The PICO question which guided the literature search did not yield information concerning denosumab that could be included. Therefore, no recommendation regarding denosumab was included in the final Guideline.
It should be noted that there are other studies that did not meet the strict scope for inclusion that examine several established BMA options for prevention of SREs in patients with multiple myeloma and metastatic carcinoma. Pamidronate 90mg IV every 3-4 weeks was found to reduce SRE's compared to placebo (Lipton, 2000; Hortobagyi, 1998). ZA 4mg IV was found to be superior to clodronic acid (Morgan, 2013). Denosumab was found to reduce risk of SREs, relative to ZA, in multiple tumor types (Lipton, 2012). Adverse event profiles differ; denosumab was associated with higher rates of hypocalcemia, while zoledronic acid was associated with acute phase reactions and renal toxicity more often. Jaw osteonecrosis rates were similar. Studies evaluating longer dosing intervals are only available for ZA, not pamidronate or denosumab.
There are few studies designed to address the question of duration of treatment with BMA's. One study in multiple myeloma patients compared ZA treatment for 4 years to 2 years, and longer treatment was associated with lower SRE rates, with similar adverse events (Aviles, 2017). Duration of treatment in a majority of the other BMA studies ranges from 1 to 3 years.Further discussion on the use of BMAs in multiple myeloma can be found in the updated American Society of Clinical Oncology (ASCO) CPG on the Role of Bone-Modifying Agents in Multiple Myeloma (Anderson, 2018).
Future research may further explore questions of treatment duration and decreased dosing frequency, with regard to not only SRE's, but also cost effectiveness and quality of life.
BMAs for Various Diagnosis
In the absence of reliable evidence, it is the opinion of the workgroup that BMAs should be considered in patients with metastatic carcinoma or multiple myeloma with bone lesions at risk for fracture regardless of tumor histology.
There is a single low-quality study by Abdel-Rahman (2018) that assessed tumor histology as a prognostic feature of skeletal related event outcomes in patients with advanced cancer and bone metastases treated with either denosumab or zoledronic acid in a clinical trial. The aforementioned study found that patients with non-small cell lung cancer had a shorter time to first skeletal related event than patients with other cancers. Despite the lack of evidence-based recommendations for this topic, clinicians should consider the use of BMAs regardless of tumor histology in patients with metastatic carcinoma or multiple myeloma with bone lesions at risk for fracture.
Although not meeting criteria for inclusion in analysis for this question in particular, there is evidence that specific BMAs may be favored by histology type. In multiple myeloma, zoledronic acid has been found to be superior to clodronate (Morgan 2011,2013) for skeletal related events, progression free survival (PFS) and overall survival (OS), whereas denosumab is non-inferior to zoledronic acid (Raje 2018) for skeletal related events and OS, but associated with a longer PFS. In patients with breast cancer, denosumab is shown to be superior to zoledronic acid in relation to reduced rates of skeletal related events, prolonged time to first skeletal related event and improved quality of life measures (Martin 2012, Stopeck 2010). In patients with prostate cancer, denosumab prolonged time to first skeletal related event compared to zoledronic acid (Fizazi, 2011). In a sub-study analysis, denosumab compared to zoledronic acid was associated with improved OS in patients with non-small cell lung cancer and bone metastases (Scagliotti, 2012)
Benefits/Harms of Implementation
The benefits of decreased fracture rates, avoiding surgical intervention and associated pain, reduction in other skeletal related events and improved survival (in some patients according to histologic type) weighed against the harms of osteonecrosis of the jaw and hypocalcemia, favor the use of BMAs in these populations. It is important for clinicians to be aware that renal insufficiency is observed more commonly for zoledronic compared to denosumab, whereas hypocalcemia is more frequently observed with denosumab.
Imaging Findings and Atypical Fractures
In the absence of reliable evidence, it is the opinion of the workgroup that imaging findings of lateral cortical thickening may be associated with increased atypical femur fracture risk.
Atypical femur fracture is a well-recognized complication of long-term administration of bone targeted agents. These fractures have consistent radiographic features, typically starting as thickening of the lateral cortex ("beaking") in the subtrochanteric or diaphyseal femur. A transverse radiolucency that develops through the lateral thickening is concerning for impending fracture and is sometimes referred to as the "dreaded black line" (Kim, 2014). If these signs are undetected, the patient may progress onto a transverse or oblique fracture (Shane, 2014). In 70% of patients, fracture is preceded by prodromal thigh pain (Dell, 2018).
Atypical femur fractures are believed to be stress or insufficiency reactions, possibly exacerbated by reduced remodeling at the fracture site due to the action of bisphosphonates (Shane, 2014). Multiple epidemiological studies have documented increased incidence of subtrochanteric fractures
The incidence of atypical femur fractures in one large population study was 55 per 100,000 person-years, compared with 1 per 100,000 person-years in bisphosphonate-naive patients (Van De Laarschot, 2017). However, it is important to remember that an estimated 162 osteoporosis- related fractures are prevented for every 1 AFF that may be associated with treatment with an antiresorptive medication (Van De Laarschot, 2017). Asians may be up to 8 times more at risk for AFF than whites (Dell, 2018). Concurrent use of glucocorticoids is associated with increased risk of AFF (Shane, 2014), which may be relevant to patients being treated for multiple myeloma.
Dual-Energy X-ray Absorptiometry Images (DEXA) scanning, used routinely in surveillance for osteoporosis, has been shown to be effective in screening for lateral cortical thickening (Kim, 2014). DEXA scanning also requires significantly less radiation exposure than routine radiographs (Van De Laarschot, 2017). One retrospective review noted a 40% of AFFs occur in the diaphysis (Unnanuntana, 2012), therefore it is important that screening DEXA scans are extended to include the diaphysis (Unnanuntana, 2012).
If a patient suffers an atypical femur fracture, stopping bisphosphonates exposure can reduce contralateral fracture, which is otherwise ~25%. There is some evidence that treatment benefit from bisphosphonates reduces after 5 years, while risk of AFF increases from 1.78/100k/year to 113/100k/year with exposure >8 years (Dell, 2018).
Patients with symptomatic lateral cortex thickening, medial callus formation or transverse lucency should undergo prophylactic intramedullary nailing (Shane, 2014). In the case of completed fractures, external rotation of the intramedullary nail during insertion can reduce the risk of malreduction of the bowed femur and accelerate fracture union (Park, 2017).
While many of the risk factors for atypical femur fractures have been described, a validated risk calculator and/or clinical pathway to guide physicians would be helpful. As of yet, there is no evidence on which to base guidance for how long patients with AFFs or radiographic signs concerning for AFF should go on a drug "holiday". The sharply increased risk of AFF in Asians and case reports of symptomatic and/or radiographic improvement following treatment with teriparatide may merit further investigation.
Efficacy of Radiation Therapy
Clinicians should consider the use of radiation therapy to decrease the rate of femur fractures in patients with metastatic carcinoma or multiple myeloma lesions who are deemed at increased risk based on the combination of imaging findings and lesion-related pain.
One observational study of moderate quality (Oh E. et al. 2017) among patients with metastatic lung cancer indicates a higher risk of fracture among patients with femoral metastases not treated with radiation therapy, as compared to those treated with radiation therapy. Other risk factors included lytic femur metastasis morphology and female gender. Though these data are limited to femur metastases from lung cancer and there is no randomized evidence to guide practice, the evidence and related recommendation was considered moderate strength given the high morbidity of femur fractures and the low morbidity of radiation therapy to the femur.
This recommendation addresses the question of whether radiation by itself can reduce the risk of fracture. It is not intended to alter current clinical practice wherein patients who are felt to be at high risk of pathologic fracture first undergo prophylactic stabilization.
Future research should address which femur metastases are most at risk for fracture, and hence further define when radiation therapy is required. Patients who suffer local recurrence of tumor within the femur after radiation therapy appear to have an increased risk of lesion-related pain, fracture, and need for surgical intervention. Further research is needed to accurately identify specific populations of patients who are at increased risk of tumor recurrence within the femur after radiation therapy, and to determine the risks and benefits associated with any interventions that are intended to reduce these risks.
Radiation Therapy and Prophylactic Femur Stabilization
In the absence of reliable evidence, it is the opinion of the workgroup that clinicians may consider the use of radiation therapy in patients undergoing prophylactic femur stabilization to reduce pain, improve functional status, and reduce the need for further intervention.
One small, retrospective study (Townsend, 1995) demonstrated that patients receiving postoperative radiation therapy following prophylactic stabilization for femur metastases had less pain, better limb function, less need of revision surgery, and better overall survival. The small, retrospective nature of this study, hampered by selection factors, renders this low-quality evidence. However, given the low morbidity of postoperative radiation therapy, and the importance of improving quality of life outcomes and reducing the need for further surgical interventions, the use of radiation may be considered for patients with metastases to the femur requiring prophylactic stabilization.
Radiation Therapy after Resection and Reconstruction
In the absence of reliable evidence, it is the opinion of the workgroup that radiation therapy may be considered after resection and reconstruction to reduce pain, improve functional status, and reduce the need for further intervention in patients with residual tumor, or those at increased risk of tumor recurrence in the setting of metastatic carcinoma or multiple myeloma of the femur.
No investigations directly compare the impact of radiation therapy after resection and reconstruction, limiting the statements that can be made about whether radiation therapy can improve outcomes in the setting of residual disease or when there is an increased risk of tumor recurrence. However, given that radiation therapy of the femur is generally well-tolerated and residual/recurrent tumor of the femur can remain/become symptomatic, the potential benefits may be felt to outweigh the harms in select patients.
Multi-Fraction Radiation Treatment
Clinicians should consider the use of multi-fraction in lieu of single fraction radiation treatment to reduce the risk of fracture in patients with metastatic carcinoma in the femur.
One randomized study of moderate quality (Van Der Linden 2003) demonstrated that multi- fraction radiation therapy was associated with a lower risk of femoral fracture compared to single-fraction radiation therapy. In the absence of other randomized data, the strength of this recommendation was
Estimating Survival and Reconstruction Method
In the absence of reliable evidence, it is the opinion of the workgroup that surgeons utilize a validated method of estimating survival of the patient in choosing the method of reconstruction. Longer survival estimates may justify more durable reconstruction methods such as arthroplasty, if clinically appropriate.
Metastatic bone disease presents unique surgical challenges within a very diverse patient population. Rather than base treatment decisions on radiographs alone, surgeons may consider the use a validated means to estimate survival such as the Tokuhashi method (Tokuhashi 2005), the PATHFx tool, available at www.pathfx.org (Ogura, 2017) or the Global Spine Tumour Study Group at www.spinemet.com. Doing so helps ensure other characteristics such as oncologic diagnosis, extent of metastases, hemoglobin, and performance status are considered when deciding on a treatment course. In general, short survival estimates (1-6 months) justify less invasive and less durable approaches, such as intramedullary nails, or less commonly, other internal fixation devices. Similarly, patients with longer estimates (>6 months) require more durable solutions such as endoprostheses, whenever possible. Patients with very short survival estimates of approximately one month may not be candidates for prophylactic fixation but may benefit from minimally or non-invasive nterventions such as radiotherapy, cryotherapy, orradio-frequency ablation for adequate pain relief (Meares, 2019; Korian, 2018). However, arthroplasty may still be indicated in patients with short survival time for palliation in certainclinical scenarios, for example fractured femoral neck.
Long Stem Hemiarthroplasty
In the absence of reliable evidence, it is the opinion of the workgroup that when treating a femoral neck fracture with hemiarthroplasty, use of a long stem can be associated with increased intra-operative and post-operative complications and should only be used in patients with additional lesions in the femur.
None of the included investigations in this clinical practice guideline directly compare short versus long stem hemiarthroplasty in this population. This limits the statement that can be made recommending one option over another. However, some evidence does exist demonstrating increased complication rates with the use of long stem cemented arthroplasty. Intraoperative hypotension and significant cardiopulmonary events including death have been documented in numerus studies (Herrenbruck 2002, Houdek 2017, Xing 2013), while other studies have demonstrated that long stem cemented implants are overall a relatively safe option if performed appropriately (Price 2013, Peterson 2017). The theoretical benefit of a long stem implant is to protect the majority of the femur from fracture in the setting of disease progression. However, some evidence does exist showing that reoperation rates in general are very low in this population and no different has been appreciated based on the length of the stem chosen (Xing 2013).
Benefits/Harms of Implementation
There may also be some additional cost involved with performing a long stem technique due to implant cost, operative time, and complication rates. Although both short stem and long stem options are at times acceptable and feasible, we believe that the potential risk involved with the long stem option is not warranted without obvious, symptomatic, concerning lesions more distal in the femur.
Future direct comparisons of short and long stem options in a randomized trial would help to clarify the question. Additional studies investigating the use of short versus long stems in patients with distal disease in the femur would help to identify which patients would benefit from short versus long stem hemiarthroplasty procedures.
In the absence of reliable evidence, it is the opinion of the workgroup that there is no advantage to routine use of cephalomedullary nails for diaphyseal metastatic lesions as there does not appear to be a high frequency of new femoral neck lesions following intramedullary nailing.
The lack of relevant and high-quality evidence regarding this topic led to a consensus level recommendation. Though it did not meet the strict inclusion criteria for this CPG, one study examined the occurrence of femoral neck metastases posterior to intramedullary nail fixation performed for a femoral diaphyseal metastatic lesion (Moon, 2015). The study reported no new femoral neck secondary lesions occurring subsequent to the aforementioned procedure.
Benefits/Harms of Implementation
Failure to diagnose a femoral neck lesion prior to implanting an intramedullary nail, increases the risks of adverse outcomes such as implant failure and the need for additional surgery. Efforts should be made to assess the entire bone length prior to decision making. Other risks are equal to those of any intramedullary nailing procedure in a cancer patient, which should be assessed on an individual basis.
The benefits of implementing this recommendation, when correctly indicated, will be the decrease in surgical time and radiation exposure to the surgeon and operating room personnel. This has implications on cost savings to society.
Further studies would be beneficial with appropriately randomized samples, power, and follow up times, examining the intramedullary nail revision rate due to the occurrence of new femoral neck lesions in the setting of metastatic disease and pathological fractures due to diaphyseal lesions.
Clinicians may consider arthroplasty to improve patient function and decrease the need for post-operative radiation therapy in patients with pathologic fractures from metastatic carcinoma in the femur.
Arthroplasty procedures carry greater potential morbidity and higher healthcare costs than internal fixation. However, these procedures may be indicated in select patients with longer expected survival and higher performance status. Four low-quality studies (Gao, H., 2016, Sarahrudi, K., 2009, Tsuda, Y., 2016, and Zacherl, M., 2011) reported comparative outcomes between arthroplasty and internal fixation for pathologic fractures of the proximal femur. Results from these studies indicate that the benefits of arthroplasty include improved function as determined by Harris Hip Scores, and a decreased need for adjuvant radiotherapy for disease control. Surgical management with both arthroplasty and internal fixation provides immediate stability to the femur and the opportunity for early post-operative mobility. The differences in outcomes are small and therefore both treatment options are reasonable.
Benefits/Harms of Implementation
Despite the benefits of arthroplasty, these procedures carry a higher risk of post-operative complications such as dislocation.
Given the increased cost of arthroplasty and the small increased risk for dislocation, the benefits of improved function and less need for radiation may not offset the cost and risks for all patients. Future studies can determine which patient characteristics are most likely to result in benefit from arthroplasty procedures in this population.
Guideline Work Group:
- Felasfa Wodajo, MD, MTST Co-Chair
- Patrick Getty, MD, MSTS Co-Chair
- Josh Petit, MD, ASTRO Co-Chair
- John Charlson, MD, ASCO, Co-Chair
- Tracy Balboni, MD, MPH
- Alan Blank, MD, MS
- Ana Cecilia Belzarena, MD
- Jonathan A. Forsberg, MD, PhD
- Michelle Ghert, MD, FRCSC
- Richard W. RIcholas, MD
- Frank Passero, MD
- Yolanda Tseng, MD, MPhil
- Kyle Mullen, MPH, Manager, AAOS Clincal Quality and Value Department
- Nicole Nelson, MPH, Lead Research Analyst, AAOS Clinical Quality and Value Department
- Anne Woznica, MLIS, AHIP, Medical Research Librarian, AAOS Clinical Quality and Value Department
- Tyler Verity, BA, Medical Research Librarian, AAOS Clinical Quality and Value Department