Background: Fixation of complex fractures of the proximal humerus is challenging. Fixed angle plates have been shown to give good results in younger patients with good bone quality, however failure of screws to maintain fixation in older patients remains a problem [1]. It has been shown that the bone quality within the humeral head sharply declines with age leaving a large bone void [2]. We propose that filling this bone void with a synthetic bone graft will improve screw purchase, and reduce the likelihood of construct failure.
Aims: We aim to use CT based finite element analysis to examine the effect of augmenting plate fixation with synthetic bone graft in the presence of a poor bone stock.
Methods: A computer tomography (CT) scan of an intact cadaveric fresh frozen humerus from a 78 year old male was obtained. The CT Hounsfield units were calibrated using a water phantom. Both the external contour and internal structure of the hummers were accurately defined using the threshold method. Using proprietary software [Simpleware Simpleware Ltd., Exeter], Boolean subtraction was employed to simulate an anatomically reduced four-part proximal humerus fracture with a representative bone void within the humeral head. A digital representation of a fixed angle proximal humerus plate was created and located so as to fix the fracture. The geometry of the plate fixed four-part fracture was then used to create a hexa-hederal dominant finite element mesh with over six hundred thousand elements created. Linear elastic properties were assigned to each element within the mesh representing bone using established relationships based on local Hounsfield number in the original CT scan [3]. The model was imported into the finite element preprocessor [Abaqus CAE, Simulia Inc, USA]. Contact interactions between the bone fragments, implant and bone graft substitute were defined. A pressure load was applied to the articular fragment to simulate maximum physiological joint reaction forces on the proximal humerus [4,5]. Simulations were run on the facilities at the Irish Center for High End Computing (ICHEC) The effect of adding synthetic bone graft to fixation with a fixed angle device was studied.
Results: In all models the peak pressures were along the lateral cortex and at the implant bone interface. This agrees with common clinical modes of failure being lateral collapse, valgus impaction and cutout of screws into the glenoid fossa. Finite element models where the simulated bone void was filled with bone graft substitute showed 60% reduction in the bearing pressures at the implant/bone interface.
Conclusions: Our results suggest that augmenting plate based fixation of complex proximal humerus fractures with synthetic bone graft results in a more robust construct. It a clinical setting this is likely to result a lower incidence of failure of fixation and subsequent revision surgery.
