Chapter 45 What Is the Appropriate Timing of Prophylactic Stabilization of Osseous Metastases?
Metastatic bone disease continues to challenge even the most experienced orthopedic surgeon. The increased survival rate of patients with carcinoma has led to an increasing population of patients with osseous metastases. Bone is the third most common site of metastasis after lung and liver.1 In up to 25% of patients, metastatic bone disease is the first presentation of carcinoma. Metastasis to bone is common, with 50% of newly diagnosed cancers eventually spreading to bone.2 The most common primary tumors that metastasize to bone are prostate (36%), breast (32%), and lung (14%).2 The most common sites for metastasis to bone are vertebrae, pelvis, ribs, femur, and skull. In as many as 10% of patients with metastases, the primary site is never found. Survival after diagnosis of bone metastasis is related to the primary bone tumor: breast, 34 months; prostate, 24 months; cervix, 18 months; colorectal, 13 months; lung, less than 12 months; and melanoma, 3 months.2 A variation exists in survival related to types and grades of primary tumor.
Plain radiographs of the involved limb in at least two planes form the initial investigation. The metastatic lesion must destroy 30% to 50% of bone and reach a size of 1 cm to be seen on plain radiographs. The characteristic features seen include radiolucent (osteolytic) lesion, although the lesion can be osteoblastic, radiodense, or mixed; minimal periosteal reaction; epicenter in the intramedullary canal; and intracortical/juxtacortical locations are rare. A bone scan demonstrates metastatic lesions on average 3 months before plain radiographs, and it can detect a lesion as small as 2 mm. McNeil3 reviewed 273 patients with known primaries and a positive bone scan. He found that 55% had actual metastases, whereas the other 45% had other reasons: trauma in 25%, infection in 10%, and miscellaneous in 10% of patients. Overall, the diagnosis of metastatic bone disease can be made in 66% of patients with a careful history and physical examination, chest radiograph, and computed tomography (CT) of the chest. An additional 13% can be diagnosed with a CT scan of the abdomen. A biopsy of the skeletal lesion adds a further 8% and allows confirmation in 28% of patients.
EVIDENCE
Ward and colleagues4 compared the results of 97 impending fractures compared with 85 complete fractures. In the impending group, there was less blood loss (438 vs. 636 mL), shorter hospital stays (7 vs. 11 days), greater likelihood of discharge home as opposed to an extended care facility (79% vs. 56%), and a greater likelihood of resuming support-free ambulation (35% vs. 12%).4
It has been recommended that femoral metastases can be treated without surgery if they are small (<2.5 cm), less than 50% of the diameter of the cortex, and are in a low-risk location (i.e., not in the pertrochanteric region of the femur)5,6 (Level V). Often, a CT scan of the affected area is valuable in determining the exact size and location of the lesion. These recommendations are controversial, serving only as guidelines, and treatment should be individualized.
Fidler7 examined the concept of greater than 50% cortical diameter as a criterion for operative intervention (Level II). The fracture risk was low (2%) when the lesion size was less than 50% of the diameter. The fracture risk was high (80%) when the lesion was greater than 75% of the diameter. Finally, when the lesion was 50% to 75% of the diameter, there was a 61% incidence rate of fracture.
Beals and coauthors8 reviewed 338 patients with 94 femoral metastases; 8 of 19 with a pathologic fracture had a lesion smaller than 2.5 cm on radiographs (Level II). This was a false-negative rate of 42% for the concept of 2.5 cm being a guideline.
Finally, Keene and coworkers9 could not use the criteria to predict which pathologic fractures would occur in patients with metastatic breast carcinoma to the proximal femur (Level II). The ranges of sizes and percentages of the cortical involvement on plain radiographs were the same for patients with and without fractures.
Several biomechanical studies have been completed to resolve the issue of pathologic fracture risk. Sprujit and researchers10 used finite element models from quantitative CT scans of cadaveric femora with lesions created in the subtrochanteric region. The model was able to successfully predict torsional failure in 82% of cases. Lee also developed a model using quantitative CT combined with engineering beam analysis to predict the load-carrying capacity of femurs with metastatic defects.10a
Roth and colleagues11 used biomechanically based models to predict metastatic burst fracture risk. They examined 92 vertebrae with osteolytic spinal metastases. Specific burst fracture risk was calculated incorporating load-bearing capacity (tumor volume, trabecular bone density, disc quality, and pedicle involvement) and load-bearing requirement (pressure load, loading rate). The models were able to predict fracture in patients at high risk and predict stability in patients at low risk.
Tschirhart and investigators12