Team Approach: The Treatment of Metastatic Tumors of the Femoral Diaphysis
- Michael B. O’Sullivan, MD1,
- Debasmita Saha, MD1,
- Jessica M. Clement, MD1,
- Robert J. Dowsett, MD1,
- Rafael A. Pacheco, MD1 and
- Tessa Balach, MD2,a
- 1Department of Orthopaedic Surgery (M.B.O.), Division of Hematology–Oncology (D.S. and J.M.C.), Division of Radiation Oncology (R.J.D.), and Department of Diagnostic Imaging and Therapeutics (R.A.P.), University of Connecticut Health, Farmington, Connecticut
- 2Department of Orthopaedic Surgery, The University of Chicago, Chicago, Illinois
- aE-mail address for T. Balach:
➢ The differential diagnosis of an aggressive bone lesion includes metastatic disease, multiple myeloma, lymphoma, and primary sarcoma of bone. Evaluation includes radiographs of the entire bone; laboratory tests; computed tomography (CT) scanning of the chest, abdomen, and pelvis; bone scintigraphy; and biopsy.
➢ Except in rare circumstances, the treatment of skeletal metastasis is palliative and the goals of care center around pain relief and the maintenance of function.
➢ Nonoperative interventions include chemotherapy, bone-modulating agents such as bisphosphonates and denosumab, radiation therapy, and ablation with cementoplasty.
➢ When prophylactic operative stabilization is indicated to prevent pathological fracture, a cephalomedullary nail is preferred for femoral diaphyseal lesions. Postoperative external-beam radiation is indicated for local disease control.
➢ High-quality treatment of these patients relies on the close coordination of multiple different specialists.
The differential diagnosis for an adult patient presenting with an aggressive-appearing bone lesion includes metastatic disease, multiple myeloma, lymphoma, chondrosarcoma, osteosarcoma, non-neoplastic conditions such as infection, and metabolic bone diseases such as Paget disease of bone1. An accurate diagnosis is critical as it will guide the treatment and goals of care. The true prevalence of osseous metastasis is unknown. In the United States alone, estimates have indicated that 280,000 to 330,000 patients are living with osseous metastasis2,3. Patients with metastatic disease substantially outnumber those with primary malignant bone neoplasms, with an estimated 3,300 patients being diagnosed with such lesions per year4. It has been estimated that 80% of skeletal metastases are secondary to breast, prostate, lung, kidney, and thyroid adenocarcinomas, with the femur and spine representing the 2 most common sites of involvement5-7.
A 60-year-old woman presents after referral by her primary-care physician with a 2-week history of atraumatic left thigh pain that has been unresponsive to nonsteroidal anti-inflammatory drug (NSAID) treatment. The medical history includes coronary artery disease, hypertension, and chronic obstructive pulmonary disease; she has no known cancer history. Previous colonoscopies and mammograms have been negative. She has a 30-pack-year history of smoking. She reports no constitutional symptoms, no worsening of shortness of breath, and no recent weight loss. On physical examination, the lower extremity is neurovascularly intact, with tenderness to palpation of the thigh. Although she has painless passive range of motion of the hip and knee, she walks with an antalgic gait. Radiographs demonstrate an aggressive, destructive, cortically based radiolucent lesion of the middle part of the femoral diaphysis (Fig. 1).
Evaluation of a Metastatic Bone Lesion
History and Physical Examination
The initial evaluation of a patient presenting with an aggressive-appearing bone lesion involves a history and physical examination. Pain is the most common presenting symptom of patients with metastatic lesions8,9. The surgeon should investigate any history of cancer, the results of screening tests (e.g., mammography, colonoscopy, prostate examinations), and risk factors. In the study by Rougraff et al., a medical history and physical examination focusing on the breast, prostate, and thyroid ultimately identified the source of malignancy in only 3 of 40 patients presenting with an osseous lesion without a previous cancer diagnosis10. Additional evaluation is necessary to establish a diagnosis.
Routine laboratory tests, including a complete blood-cell count, comprehensive metabolic panel, erythrocyte sedimentation rate (ESR) test, C-reactive protein (CRP) test, and serum protein electrophoresis (SPEP), should be ordered by the surgeon10. The results of these tests are typically nonspecific, with the exception of the SPEP, which has been reported to have a sensitivity of 71%, a specificity of 83%, and a negative predictive value of 94% for the diagnosis of multiple myeloma in the setting of a solitary radiolucent bone lesion11. While the sensitivity is low, the negative predictive value of SPEP is sufficient to rule out the disorder. Despite the nonspecificity of the remainder of the laboratory analyses, these tests provide vital information regarding the general health of the patient. Anemia occurs in approximately 56% of patients with cancer; the surgeon should be aware of this finding prior to any interventions12. In addition, patients with breast cancer, lung cancer, renal cancer, multiple myeloma, and lymphoma may exhibit varying degrees of hypercalcemia13.
A radiographic work-up is initiated by the surgeon with consultation from the radiologist to evaluate the lesion and to identify both the primary site and any concurrent sites of disease. Orthogonal radiographs of the entire bone should be made to evaluate the lesion and to identify discontinuous lesions in the same bone. For the evaluation of a metastatic skeletal lesion alone, radiographs are typically sufficient1,14. However, if the lesion is difficult to visualize, a computed tomography (CT) scan may help the surgeon to guide local treatment. In addition to local imaging, a CT scan of the chest, abdomen, and pelvis with oral and intravenous contrast media should be performed to help the radiologist to identify the primary site of disease and to evaluate for visceral metastases. In a prospective study of occult skeletal metastases, CT scans of the chest, abdomen, and pelvis detected the primary site of malignancy in 28 (70%) of 40 patients10. The surgeon should order bone scintigraphy to identify additional sites of skeletal disease that may require local treatment. Increased radiotracer uptake on bone scintigraphy imaging occurs at sites of increased osteoblastic reparative activity6. In some lytic lesions, such as multiple myeloma or metastatic renal cell carcinoma, osteoclastic lysis prevails over the osteoblastic reparative process, resulting in little to no radiotracer uptake, thus yielding false-negative results15,16. In contrast, positron emission tomography (PET) utilizes a tracer that mimics glucose; as such, uptake is highest in areas of high metabolism (i.e., cancer). By directly measuring the metabolic activity of the tumor compared with that of the surrounding normal host bone, PET can be superior for the detection of some osseous metastases but may result in false-positive results1,15-18. However, given the sizable cost of PET, bone scintigraphy continues to be the first-line imaging modality to evaluate for polyostotic skeletal metastasis19.
In the study by Rougraff et al., a thorough history and physical examination, routine laboratory work-up, radiographs of the involved bone and chest, whole-body bone scintigraphy, and CT scans of the chest, abdomen, and pelvis identified the primary site of malignancy in 85% of individuals with a skeletal metastasis of unknown origin10.
Establishing a tissue diagnosis is the next step in the evaluation. However, the biopsy should not take place until the previously described staging work-up has been completed to guide biopsy planning14. Core-needle or incisional biopsies are preferred over fine-needle aspiration as the former techniques allow for a more complete histological evaluation of the tissue. Image-guided core-needle biopsy has been reported to have a 67% diagnostic yield (284 of 423) for osseous lesions20. Incisional biopsy is less prone to sampling error and can more accurately indicate the tumor grade but often requires a general anesthetic and may be associated with increased morbidity1,21,22. In addition, the consequences of a poorly planned open biopsy are considerable. The biopsy modality that is utilized should be decided on a case-by-case basis by the surgeon, who would perform the open biopsy, and the interventional radiologist, who would perform the core-needle biopsy. If a core-needle biopsy is performed, the surgeon can wait for the definitive diagnosis before proceeding with treatment. When an open biopsy is performed at the time of surgery for a presumed metastatic lesion (e.g., treatment of an impending or pathological fracture), the biopsy should be performed first. The surgeon can proceed with definitive surgery if frozen-section analysis indicates metastatic adenocarcinoma. If the frozen-section analysis is inconclusive or suggests a primary bone sarcoma, the surgeon should not attempt definitive treatment. In that scenario, the surgeon should obtain more tissue for pathological analysis and culture. Fractures can then be stabilized with a short plate, whereas impending fractures should be treated with the insertion of cement into the biopsy hole prior to wound closure. The surgeon should then await the final pathological diagnosis before attempting definitive surgery. Direct communication with the pathologist regarding the adequacy of the specimen and its handling helps to increase the likelihood of successful biopsy and accurate diagnosis.
Patients with metastatic carcinoma have a limited life expectancy. In a series of 141 patients presenting with metastatic lesions in the femoral shaft, the median survival was 9 months23. A patient presenting with osseous metastasis to the femur is typically incurable with surgery alone. As such, treatment is centered on improving the quality of life by minimizing pain and maintaining function with a variety of care modalities1. Close coordination of the multidisciplinary team for these patients is imperative as the disease burden, prognosis, and specific goals of care heavily influence the personalized treatment plan for these lesions.
Once a tissue diagnosis is made, the surgeon must determine whether surgical stabilization is warranted. In the setting of a pathological fracture through a metastatic femoral diaphyseal lesion, operative intervention is indicated. For patients presenting with an impending fracture, prophylactic internal fixation may be undertaken to increase the mechanical stability of the bone and to prevent pathological fracture. The available evidence suggests that the treatment of impending fractures is preferred because prophylactic stabilization may result in shorter operative times; may prevent pain and disability associated with fractures; may result in lower transfusion rates, earlier mobilization, shortened hospital stays, and higher rates of discharge to home; and may allow more patients to maintain support-free walking compared with those who have a pathological fracture24-26. In addition, pathological fractures have an impaired ability to heal, resulting in a high rate of delayed union or nonunion. Gainor and Buchert27 showed that osseous union was achieved in only 35% of patients. In the group of patients who survived >6 months following fracture, 74% of fractures united. Given this propensity for nonunion and delayed union, pathological fractures have been associated with a high rate of implant failure and reoperation28,29.
While prophylactic treatment is preferable, the ideal timing of surgery is controversial. Unfortunately, most of the literature about the treatment of impending fractures is based on data from retrospective reviews involving small sample sizes. Physicians frequently rely on the Harrington30 definition of impending pathological fracture or the Mirels31 classification system to determine the risk of fracture1,32. Harrington30 defined an impending fracture as destruction of ≥50% of the circumference of the cortical bone, a proximal femoral lesion measuring ≥2.5 cm in any dimension, a pathological avulsion fracture of the lesser trochanter, or persistent stress pain despite radiation. However, Van der Linden et al.33 found that a lesion size of >2.5 cm and increasing pain were not predictive of fracture. Also, Keene et al.8 found no differences between patients who sustained pathological fractures and those who did not in terms of lesion size and pain level, limiting the clinical utility of this definition. On the basis of a retrospective review of 78 metastatic lesions with 6 months of follow-up data, Mirels31 developed a classification system to assess fracture risk (Table I). According to that system, patients with a score of ≥9 should undergo prophylactic internal stabilization. Although the Mirels31 criteria demonstrated a sensitivity of 91%, which was higher than that of clinical judgment alone, they had a specificity of only 35% for predicting fracture32.
Recently, CT-based finite-element modeling (FEM) and structural rigidity analysis (CTRA) have been prospectively evaluated for fracture prediction in patients with metastatic lesions of the femur34-36. In the study by Goodheart et al., FEM accurately predicted that the force needed to cause a fracture was lower in patients who ultimately sustained pathological fractures and, under level walking conditions, demonstrated greater specificity than the Mirels31 criteria (86% versus 43%) for fracture prediction34. In the study by Damron et al., CTRA demonstrated a higher sensitivity (100% versus 66.7%) and specificity (60.6% versus 47.9%) than the Mirels31 criteria35. Nazarian et al. reported that, when CTRA was available to physicians, this analysis changed the treatment recommendations for 36 of 124 patients and predicted fracture with 100% sensitivity and 90% specificity in the population of patients who were managed nonoperatively36. While both FEM and CTRA have demonstrated promising early results, the necessary software is not widely available35. In the absence of alternatives, it is reasonable to rely on the predictive capabilities of the Mirels31 criteria and to advocate for surgical intervention when a patient with a metastatic lesion of the femur has a score of ≥9.
If stabilization is indicated for a pathological or impending fracture of the femoral diaphysis, the proposed surgical intervention should minimize both time in the hospital and time for recovery; the implants should allow for immediate weight-bearing and be durable enough to tolerate a substantial load for the remainder of the patient’s life in the absence of healing or in the presence of local disease progression; and the treatment should aim to avoid reoperation1. Therefore, long intramedullary nails, specifically, cephalomedullary nails, are the implant of choice because they allow for protection of the entire femur in the event of additional osseous destruction or metastasis and because, as load-sharing constructs, they allow for immediate weight-bearing, in contrast to a plate-and-screw construct. Traditionally, cephalomedullary nails have been preferred as they protect the femoral neck in the event that metastases develop in the future. However, Alvi and Damron37 found that only 1 of 96 patients with osseous metastasis, myeloma, or lymphoma developed an additional metastatic lesion following surgical intervention. In addition, Moon et al. recently reported no cases of subsequent femoral neck metastases in a group of patients with diaphyseal lesions who underwent treatment for a pathological or impeding fracture23. Although that series did not support the routine use of a cephalomedullary nail for prophylactic treatment of the femoral neck, the most recent radiographs were made at a median of only 5 months after surgery. With limited radiographic follow-up and a minimal increase in morbidity associated with cephalomedullary nails, additional study is required before the use of alternative nail constructs can be recommended.
Intramedullary nails are associated with the risk of embolization of marrow contents and subsequent cardiorespiratory dysfunction. Data suggest that this risk is higher with the placement of a prophylactic nail38; because the femur is a closed compartment, instrumentation can increase the intramedullary pressure, leading to embolization of medullary contents. Pressures as high as 450 mm Hg have been measured during reaming of an intact femur; this value is well over the 25-mm-Hg threshold thought to promote embolization39,40. Barwood et al.41 demonstrated that 11 of 43 patients undergoing intramedullary nailing for the treatment of femoral metastasis had intraoperative oxygen desaturation and hypotension. To decrease this risk, a vent hole can be made in the distal femoral metaphysis, thereby decreasing the maximum intramedullary pressure and the duration of time exceeding embolic thresholds by 50%40,42. Both reaming and the insertion of a cephalomedullary nail can seed the intramedullary cavity with tumor cells. Postoperatively, the radiation oncologist should manage the patient with adjuvant external-beam radiation therapy covering the entire implant to improve local disease control.
Postoperative radiation therapy has been associated with the achievement of nearly normal levels of pain and function along with a significantly decreased reoperation rate (p = 0.03)43. There are limited data to support the use of single-fraction radiation schedules in the postoperative setting. Conway et al.44 demonstrated reasonable levels of partial pain relief (70.3% versus 74.4%) and complete pain relief (20.3% versus 31.7%) for patients managed with single as compared with multiple-fraction radiation schedules. Established practice guidelines include the coverage of the surgical implants in the radiation fields, although 1 study demonstrated that incomplete coverage of the surgical implants can be considered45.
For patients with lesions that are at low risk for impending fracture (Mirels31 score, ≤8) or who are medically unable to undergo surgery, there are a variety of nonoperative interventions to treat the pain and disability associated with metastatic lesions of the femoral diaphysis.
For patients with osseous metastasis, bisphosphonate therapy should be considered by the medical oncologist. Bisphosphonates work as osteoclast inhibitors to decrease osseous resorption. Third-generation drugs, including zoledronate (Zometa), demonstrate the most potent antiresorptive activity. Denosumab (Xgeva), an inhibitor of the receptor activator of nuclear factor kappa-B ligand (RANKL), is an alternative to bisphosphonates. These agents reduce the risk of skeletally related events, including pathological fracture, spinal cord compression, malignant hypercalcemia, and the need for radiation therapy or surgery46.
In patients with solid tumors, multiple myeloma, breast cancer, or castration-resistant prostate cancer, denosumab delayed the time to the first skeletally related event by 16% to 18% as compared with zoledronate. Denosumab also has been reported to prevent treatment-induced bone loss and pathological fractures associated with androgen deprivation therapy47-49.
Although generally well tolerated, bisphosphonates and denosumab are associated with jaw osteonecrosis and hypocalcemia. Risks specific to bisphosphonates in patients with malignancy include impaired renal function, a temporary influenza-like syndrome, and a risk of atrial fibrillation/flutter and stroke. Because of their benefits, bisphosphonates or denosumab are routinely used to prevent bone-metastasis-related problems, such as fracture and bone loss.
Many patients with bone metastases have disabling pain, making analgesics an important part of their comprehensive treatment plan8,9. It is critical that pain management take place in conjunction with other treatments.
Non-opioid analgesics such as acetaminophen or NSAIDs may be suitable for the treatment of mild pain, but attention should be paid to dose-related side effects (e.g., renal and hepatic function). When these medications do not adequately control pain, opioids should be added. Opioids require dose titration, often starting with as-needed short-acting medications and progressing to the addition of longer-acting drugs with breakthrough medications as needed50.
In refractory cases, corticosteroids may be helpful in the acute setting; a daily dose of 4 to 16 mg of dexamethasone may be considered51. Other modalities for pain control include nerve blocks, a spinal cord stimulator, an implanted epidural catheter, transcatheter embolization, or a pain pump that delivers local or systemic analgesics. In the nonoperative setting, these medications are most appropriately managed by the medical oncologist or the primary-care physician, given the need for frequent dose titration. However, consultation with a palliative-care team or a pain specialist should be considered for patients who have pain that is refractory to initial attempts at analgesic therapy or who have side effects that limit the use of analgesics.
Radiation therapy can be used as monotherapy by the radiation oncologist for pain relief, preservation of bone integrity, and maintenance of lower-extremity function in patients who do not require surgical stabilization. External-beam radiation therapy is the most commonly used technique. Radiopharmaceuticals are indicated for the treatment of painful skeletal metastasis (Sr-89) and castration-resistant prostate cancer associated with painful skeletal metastasis in the absence of visceral metastatic disease (Ra-223). These agents are often used in conjunction with localized external-beam radiation therapy when effective systemic therapies are not available52.
Common dose and fraction schedules for primary radiation therapy vary from 8 Gy in a single fraction to 30 Gy in 10 fractions. The pain response to an initial course of radiation therapy is in the range of 60% to 80%52. Randomized trials also have demonstrated the effectiveness of single-fraction schedules in relieving pain53. A single-fraction schedule is particularly appealing for patients with a poor prognosis or transportation-related issues. However, there is an increased need for retreatment after the use of lower-dose, single-fraction schedules54,55. Studies supporting single-fraction schedules largely have excluded patients with complicated lesions (e.g., fracture, impending fracture, soft-tissue component)54,55. Therefore, many clinicians favor a protracted schedule for such patients. If there is incomplete pain relief or increased pain after an initial response, retreatment with radiation can lead to additional benefit. Huisman et al., in a meta-analysis, found that retreatment resulted in pain relief in 58% of 527 patients56.
A transient increase in pain during the early portion of a radiation course can occur in 30% to 40% of patients. A brief course of dexamethasone concurrent with the initiation of radiation therapy can reduce the likelihood of this pain flare51.
Although radiation therapy is effective for the treatment of a wide variety of tumor types, the primary tumor site can influence the pain response and the duration of benefit. Patients with breast cancer metastases have shown improved pain relief compared with patients with other types of tumors57; however, even relatively resistant tumors, such as renal cell carcinoma, can have durable pain response (>1 year) in 47% of patients58.
Minimally Invasive Radiologic Interventions
For metastatic lesions that do not warrant surgical stabilization, minimally invasive treatment by an interventional radiologist is an option. When assessing a tumor for these interventions, the primary goal should be overall pain reduction, especially when the pain is refractory to opioid analgesics. Recent reports have indicated that a dual intent to reduce overall pain and to provide stability to the distal aspect of the lower extremity (e.g., the tibia and fibula) with interventional therapy has merit59. The treatment of lesions of the femoral diaphysis has not proven to be as successful; rather, studies have shown that interventional treatments are an excellent option to reduce overall pain but are less effective for reducing the risk of pathological fracture59.
The primary treatments available are tumor ablation with or without cementoplasty. Although ablation alone can provide adequate pain relief, the addition of cementoplasty has been associated with more rapid and effective pain reduction60. Tumor ablation results in targeted cell death, limiting the local growth of the tumor. Early intervention is ideal to limit local tumor growth, reducing the likelihood of pathological fracture. Cementoplasty, the percutaneous application of viscous material (typically polymethylmethacrylate, although other, higher-viscosity cements have been used) to provide structural support in the bone plays a role in the stabilization of bone matrix by preventing further microfractures—a key component in overall pain reduction59,61. Although not an adequate substitute for surgical stabilization, treatment of femoral lesions with combined ablation and cementoplasty is a promising method of minimally invasive intervention when indicated (Fig. 2)59.
While many ablative techniques are available, radiofrequency ablation and cryoablation are the most studied and preferred methods of ablative treatment62,63. Tumor characteristics that dictate which ablative technique is utilized include size, vascularity, and proximity to nerves. Therefore, appropriate cross-sectional imaging is necessary to aid in pre-procedure planning. When ablation techniques are used, cementoplasty should always be considered to enhance the overall structural support during the treatment of bone metastases.
The role of combined tumor ablation with cementoplasty and radiation therapy has not been clearly established for the treatment of femoral lesions, but dual therapy can be extremely successful for pain reduction and can result in local tumor control; however, additional study is needed to identify tumors that are more susceptible to this type of dual therapy and to establish an ideal sequence of treatment64.
Regardless of the treatment modalities chosen to treat the pain and disability of the local disease, chemotherapy and, increasingly, targeted molecular agents are utilized to achieve global disease control. While certain cancers, such as lymphoma, respond rapidly to chemotherapy, others, such as renal cell carcinoma, are chemoresistant and thus chemotherapy would not be utilized. Table II lists some common treatment approaches for specific tumor subtypes. In recent years, there has been substantial growth in the field of targeted molecular agents and evolution in treatment regimens. Renal cell carcinoma, for example, is now treated with systemic antiangiogenic therapy, namely, drugs such as pazopanib, which improved the progression-free survival of patients from 4.2 to 9.2 months65. Given the rapidly evolving nature of this field, early involvement of a medical oncologist is essential to achieve adequate global disease control.
The management of patients presenting with femoral metastatic lesions depends on the close interaction of providers across multiple specialties, including orthopaedic surgery, radiology, interventional radiology, pathology, radiation oncology, medical oncology, and primary care. The most appropriate treatment interventions depend on both the wishes of the patient and the ultimate prognosis. Improving quality of life by decreasing pain and maintaining function remain the overarching goals. The input from multiple different specialists is required to achieve these goals.
After the initial screening work-up, a solitary lesion of the lung was identified on CT scanning. No additional sites of osseous metastasis were found with bone scintigraphy. An image-guided core-needle biopsy of the femoral lesion confirmed the diagnosis of adenocarcinoma of the lung. Given the location of the skeletal metastasis, the presence of functional pain, and the radiographic appearance and size of the lesion, the patient was determined to be at high risk for impending pathological fracture. The patient elected to undergo prophylactic internal stabilization to control pain and to avoid pathological fracture. She was managed with a cephalomedullary nail and, following an uneventful hospitalization, was discharged to home (Fig. 3). Postoperatively, she underwent external-beam radiation therapy for local disease control. Following this treatment, she noted considerable pain relief and was able to maintain support-free walking. Postoperatively, chemotherapy was initiated for the treatment of systemic disease.
Investigation performed at University of Connecticut Health, Farmington, Connecticut
Disclosure: The authors indicated that no external funding was received for any aspect of this work. The Disclosure of Potential Conflicts of Interest forms are provided with the online version of the article.
- Copyright © 2017 by The Journal of Bone and Joint Surgery, Incorporated