LAUSR

INTRODUCTION Atypical femoral fractures (AFF) are characterized as low-energy fractures of the femoral shaft or subtrochanteric region. Femoral geometry is known to play a role in AFF risk; it is hypothesized that high-risk geometries are associated with elevated femoral shaft strain. However, it is not well known which geometric parameters have the greatest effect on strain, or whether interaction between parameters is significant. The purpose of this study was to thoroughly quantify the relationship between femoral geometry and diaphyseal strain, using patient specific finite element (FE) modelling in concert with parametric mesh morphing. METHODS Ten FE models were generated from computed tomography (CT) images of cadaveric femora. Heterogeneous material properties were assigned based on average CT intensities at element locations and models were subject to loads and boundary conditions representing the stance phase of gait. Mesh morphing was used to manipulate 8 geometric parameters: neck shaft angle (NSA), neck version angle (NV), neck length (NL), femoral length (FL), lateral bowing angle (L.Bow), anterior bowing angle (A.Bow), shaft diameter (S.Dia), and cortical bone thickness (C·Th). A 2-Level full factorial analysis was used to explore the effect of different combinations of physiologically realistic minimum and maximum values for each parameter. Statistical analysis (Generalized Estimating Equations) was used to assess main effects and first order interactions of each parameter. RESULTS Six independent parameters and seven interaction terms had statistically significant (p<0.05) effects on peak strain and strained volume. For both measures, the greatest changes were caused by S.Dia, L.Bow, and A.Bow, and/or first order interactions involving two of these variables. CONCLUSIONS As hypothesized, a large number of geometric measures (six) and first order interactions (seven) are associated with changes in femoral shaft strain. These measures can be evaluated radiographically, which may have important implications for future studies investigating AFF risk in clinical populations.