This thesis aims to examine the biological and mechanical factors which influence
the strength of impacted morsellised bone graft. The use of synthetic materials as
bone graft enhancers was also analysed (Controlled Release Glass - Corglaes®,
Giltech Ltd, Ayr, Scotland and Tricalcium Phosphate Hydroxyapatite - TCP/HA,
Stryker Howmedica Osteonics, Berkshire, UK).
Phase I - the mechanical strength of different fresh frozen human bone graft
preparations were analysed in the laboratory, looking specifically at the effect of
particle size, washing the graft and using synthetic additives.
Phase II - mixtures of bone graft and bone graft/Corglaes®, which had been
identified as being mechanically stronger than standard bone graft, were analysed for
their biological response in an in-vivo ovine defect model, compared to controls.
Bone densitometry and histological analysis was performed.
Phase III - a mixture of bone graft and Corglaes® was compared to the current 'gold
standard' of allograft bone alone, in a simulated revision in-vivo ovine femoral hip
replacement model. Outcome measures of subsidence, micromotion on cyclical
loading and histomorphometry were performed.
Phase 1 - Fresh frozen human bone-graft behaves in a similar fashion to theoretical
predictions based on Engineering principles. These principles follow Soil Mechanics
theory and are commonly used by engineers when designing stable foundations for
roads or buildings. In general, when the spread of different particle sizes is uniform
over a given range, the material is stronger (more resistant to shear) than if the
particles are all the same size. This allowed determination of which bone mills
produce the strongest graft. These results were dependent on the degree of fluid
release on graft milling, with more fluid release when the average particle size is
reduced. Shear strength was improved for all mills after washing the morsellised
graft or by the addition of synthetic additives (Corglaes® and TCP/HA).
Phase 2 - The defect model allowed analysis of remodelling of the impacted pellets,
highlighting rapid dissolution of the Corglaes®, without a significant inflammatory
response. The model may be closer to a fracture model when the histological results
of phase III are considered.
Phase 3 - No statistically significant difference in subsidence over the implantation
period (12 weeks) or micromotion of the retrieved implant / femur composites could
be elicited between the two groups. Histological analysis revealed the distal impacted
graft to be in an isolated environment, both from biological ingress and solution
exchange. Bone graft and Corglaes® that was remodelled or resorbed after time in the
Phase II defect experiments, was little changed with time in the distal femur. The
proximal femur histologically behaved in a similar fashion to the defect experiments.
This suggests that a defect model alone is not ideal to analyse materials for impaction
• Graft strength is variable depending on the bone mill that produces it - washing
bone graft improves the strength from all bone mills tested.
• Tight compaction with smaller particles does not inhibit neovascularisation.
• Novel biomaterials by themselves were inferior mechanically and biologically.
• 50/50 mixes of allograft and Corglaes® are stronger mechanically and do not
appear to have an adverse effect on biological incorporation.
• In this sheep hemiarthroplasty model, subsidence, micromotion and
histomorphometry results better replicate the equivalent reported human results
than previous models, especially unloaded defect models in lower vertebrates.