The impaction allografting procedure for the treatment of failed hip reconstructions has shown promising results, but a number of intraoperative and postoperative concerns exist. A fundamental understanding of the biomechanical characteristics of the construct would help in providing surgical guidelines to address the issues of stem subsidence, femoral fracture, and resource usage. To this end, an in vitro model was developed to compare the structural response of the construct as a function of the material composition (i.e. 100% morsellized bone, 100% cement, and a combination of morsellized bone and cement as used in the traditional impaction allografting procedure), distal plug fixation (i.e. fixed or free distal plug), and stem surface finish (i.e. polished or polished and coated with mould release).

To replicate clinical conditions while eliminating confounding variables, the model system comprised aluminum tubing to represent the femur and an axisymmetric single- taper stem. One orthopaedic surgeon performed all procedures in a total of 25 specimens. Successive uniaxial cyclic compression was applied to the models to simulate 1-BW (i.e. body weight), 2-BW, and 3-BW loading for a total of 1000 cycles. Applied force, axial stem displacement, strain at the outside of the tubes were recorded to calculate construct stiffness, subsidence, and hoop and axial strain. At the conclusion of each test, the stems were pulled out of the tubes to gain further insight into the relationship among material properties, loading, and interfacial shear strengths.

The structural behaviour of the impaction allografting (IA) construct was closer to that of the 100%o cement (CE) construct than that of the 100% morsellized bone (MB) construct. The IA and CE constructs did not show substantial stem subsidence and hence exhibited predominantly axial strain resulting from high shear stresses at the stem-cement interface. Conversely, the MB specimens showed excessive stem subsidence and consequent increase in the hoop:​axial strain ratio with morsellized bone compression. There was no change in the structural behaviour of the specimens when a mould release-coated stem was substituted for the polished stem. With the distal plug removed, the IA and CE specimens slipped at the tube-material interface, suggesting that the distal plug or some other form of external or geometric support is required for structural integrity.