The accurate assessment of a patient's risk of fracture in the presence of a benign or metastatic skeletal lesion can help determine the optimal treatment and/or intervention plan. Unfortunately, reliable predictors of pathologic fracture have been elusive. In this study, Snyder et al. address the specific question of predicting fracture risk in children and young adults who have a benign skeletal lesion. The authors note that the traditional approaches for predicting the risk of pathologic fracture, which have been chosen on the basis of factors such as site, size, and stage of the lesion and percentage of cortical destruction as seen on plain radiographs, have met with modest success at best. In this unique clinical study, Snyder et al. hypothesize that a mechanics-of-structures approach that simultaneously accounts for spatial variations in the material properties of bone tissue and whole-bone geometry (including alterations due to the disease process) would provide a better prediction of fracture risk of benign skeletal lesions than would an approach that was based on parameters obtained from plain radiographs.
The authors have been conducting in vitro studies in support of this clinical approach for some time. Snyder and colleagues recently reported on the load-bearing capacity of bone cylinders (they cleverly obtained large, homogeneous specimens by using whale vertebral trabeculae) into which simulated lytic defects were introduced1. They found that the load-bearing ability of the specimens was directly proportional to the axial, bending, and torsional rigidities of the cylinders at their weakest cross-section. Rigidity is the resistance to deformation of a structure (e.g., whole bone) under different loading modalities and is the product of a material property and a geometrical property. For example, axial rigidity is the product of the elastic modulus, E, of the bone tissue (a measure of the intrinsic stiffness of the tissue) and the cross-sectional area of the bone. (A tutorial on the concept of structural rigidity and its application to whole-bone behavior is nicely provided by the authors in this article and can also be found in more detail elsewhere2.)
In their current paper, to test their hypothesis, Snyder et al. evaluated thirty-six patients who had benign osteolytic lesions; eighteen patients had a pathologic fracture through the lesion, and eighteen patients had no fracture at the time of the two-year follow-up. The patients all underwent clinical, radiographic, and quantitative computed tomographic evaluations. Quantitative computed tomography was conducted on both the affected bone and the contralateral bone (as a control), and this information was used to extract material and geometric properties for the structural analyses. Importantly, material properties were mapped spatially and were assigned with use of experimentally established empirical relationships between elastic modulus and the apparent density of bone tissue. The ratio of the rigidity of the affected bone to the contralateral bone was then determined to provide a measure of the loss in load-carrying capacity of the affected bone. The structural rigidity measurements were considered in light of the probability of fracture. Plain radiographs were used to assign fracture risk on the basis of length, width, or extent of involvement of the cortex. Among other comparisons between patient cohorts, sensitivity (correct detection of patients with a fracture) and specificity (correct detection of patients without a fracture) of the quantitative computed tomographic and radiographic parameters were determined.
The authors found that standard radiographic criteria based on the size of the defect were weak predictors of fracture risk, both with regard to sensitivity (28% to 83%) and specificity (6% to 78%). In sharp contrast, all biomechanical parameters that were obtained with use of quantitative computed tomography were 100% sensitive and were 44% to 89% specific. Further, when the bending and torsional rigidity measures were combined, specificity increased to an impressive 94%. This result is consistent with the observation that all the fractures occurred either in bending or torsion; put another way, failure in axial compression is an unusual mode of clinical failure. Notably, the use of bone mineral content alone as a predictor of fracture risk was 100% sensitive, but it was associated with poor specificity (44%).
This is a well-conducted, important clinical study because it demonstrates and validates the strong predictive capability of a biomechanical assay (structural rigidity) wherein material properties and geometry are accounted for together rather than separately. The approach extends the early observations of others (e.g., Burstein et al.3) that engineering mechanics can be utilized to describe the resistance to deformation and fracture of a whole bone in the presence of a defect. However, this is the first patient-based study to demonstrate that such an approach can be conducted noninvasively; it is also the first to validate this approach for the prediction of pathologic fractures. In addition, as the authors note, the analysis can also be used as a biomechanical assay by which to assess alteration in the risk of bone fracture following treatment. Notably, this is a patient-specific methodology; it does not require knowledge of the outcomes of a large patient population.
A limitation of this study is that the analytic tools necessary to perform this technique are not currently available in a clinical setting outside of the authors' institution. In addition, quantitative computed tomography exposes the patient to large doses of radiation, which would discourage repeat scans. Finally, the approach relies on quantitative computed tomographic analysis of a normal contralateral limb, which may not be possible if that limb also has undergone pathologic changes. These limitations are acknowledged by the authors, and all can be addressed. Efforts are underway by the authors to make the analyses more accessible to clinicians through existing image-analysis software packages. In the meantime, the authors are willing to analyze quantitative computed tomographic images that are sent to them. Radiation exposure can be reduced or eliminated through the use of new multidetector helical computed tomography scanners or by obtaining the structural information via magnetic resonance imaging (an approach being developed by the authors)4. In the presence of an involved contralateral limb, the risk of bone fracture can be assessed by calculating the failure load of the bone and then comparing it to the load associated with a particular activity. The extension of this approach to the evaluation of metastatic bone lesions is a logical next application. In fact, a related in vivo study to assess the fracture risk of metastatic bone lesions is in progress5.
In summary, the authors have developed a powerful and robust noninvasive tool, based on fundamental principles of engineering mechanics, that can be used to reliably identify patients at risk of sustaining a fracture through benign osteolytic lesions. It is easy to imagine that this approach could also be extended to prediction of fracture in bones that are globally at risk (e.g., osteoporotic bones). In the future, this article will undoubtedly be widely referenced for its impact on the approach to evaluation of patients at risk for pathologic fractures.
*The author did not receive grants or outside funding in support of her research for or preparation of this manuscript. She did not receive payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the author is affiliated or associated.
1. Hong J, Cabe GD, Tedrow JR, Hipp JA, Snyder BD. Failure of trabecular bone with simulated lytic defects can be predicted non-invasively by structural analysis. J Orthop Res. 2004;22:479-86.
2. Burstein AH, Wright TM. Fundamentals of Orthopaedic Biomechanics. Baltimore: Williams and Wilkins; 1994.
3. Burstein AH, Currey JD, Frankel VH, Heiple KG, Lunseth P, Vessely JC. Bone strength: The effect of screw holes. J Bone Joint Surg Am. 1972;54:1143-56.
4. Hong J, Hipp JA, Mulkern RV, Jaramillo D, Snyder BD. Magnetic resonance imaging measurements of bone density and cross-sectional geometry. Calcif Tissue Int. 2000;66:74-8.
5. Cordio MA, Snyder BD, Wilson SB, Kwak SD, Zurakowski D, Parker L. Noninvasive prediction of fracture in patients with metastatic breast cancer to the spine. Oncology. 2003;17(S3):29.