Importance of Metastatic Lesion Location and Subject‐Specificity in the Determination of Femoral Fracture Risk

Evaluating the risk of pathological fracture in patients with bone metastasis continues to be a challenge. Clinicians use scoring systems such as Mirels', which are known to be unreliable. Patient-specific finite element (FE) analyses have been shown to be more effective than empirical clinical guidelines. While patient-specific FE models are valuable, they do not provide trends on fracture risk with respect to lesion locations that may apply across patients. Also, undertaking scans and conducting simulations for every patient is not practicable. Current knowledge is limited regarding the effect of lesion location in the femur and influence of patient-specific factors on fracture risk. We developed an automated system that generates synthetic spherical lesions of uniform size, systematically shifts their location by 1 mm, and evaluates the resulting effects on the mechanical response to loading applied at the femoral head aligned to the mechanical axis. To evaluate the importance of subject-specificity, models developed from CT scans of three cadaveric femurs and a generic Sawbones model were analysed for their mechanical behaviour for similar variation in lesion location. We found that lesion location plays an extremely important role in the determination of fracture risk, and that the trends associated with location are similar across subjects. Lesions in the femoral diaphysis with no cortical involvement have no distinguishable impact, while loss of cortical bone and their location from medial to lateral and along the shaft (proximal-, mid- and distal-diaphysis) have a key role in predicting potential fracture. We also found that normalised loss of stiffness when the lesion is on the medial side is almost two times that on the lateral side, as long as there is some cortical involvement in the proximal- and mid-diaphysis.