Introduction: Unicompartmental knee replacements (UKRs) are an alternative to total knee replacements (TKRs) for treating isolated medial compartment knee osteoarthritis. However, revision rates are consistently higher than for TKR and UKRs are commonly revised for “unexplained” pain, a possible cause of which is elevated proximal tibial bone strain. The influence of implant design on this strain has not been previously investigated. Aims: The aims of this thesis are to determine the effect of medial UKR tibial component design on proximal tibial strain and ongoing pain. Methods: A retrospective clinical cohort study was performed comparing patient reported outcome and implant survival of a metal backed mobile bearing UKR implant (n=289) and an all-polyethylene (AP) fixed bearing UKR implant (n=111) with minimum 5 year follow up. A method of digital radiological densitometry, the greyscale ratio b (GSRb), was developed, validated and applied to plain radiographs to measure changes in bone density over 5 years in both the metal backed (n=173) and all-polyethylene (n=72) UKR patients. A finite element model (FEM) was validated against previous mechanical testing data and was used to analyse the effect of metal backing and implant thickness on proximal tibial cancellous bone strain in fixed bearing UKR implants. Results: There were no significant differences in patient reported outcomes between implants throughout follow up. Ten year all cause survival was 90.2 (95%CI 86-94) for the metal backed implant and 79.9 (60.7 to 99) for the all-polyethylene. Revision for unexplained pain was significantly greater in the AP implant where revisions were performed significantly earlier. Overall, the mean GSRb reduced following medial UKR with no difference between implants. In those patients where GSRb increased, patient reported outcomes were worse with an association with ongoing pain. A finite element model was successfully validated using acoustic emission and digital image correlation data. This model confirmed that the volume of cancellous bone exposed to compressive and tensile strains in excess of 3000 (pathological overloading) and 7000 (fracture) microstrain were higher in the AP implants, as were peak tensile and compressive strains. Varying polyethylene insert thickness did not affect these strain parameters in the metal backed implant, but varying polyethylene thickness in the AP implants had significant effects at all loads with elevated strains in thinner implants. Increasing the AP thickness to 10mm did not reduce strains to the levels found under metal backed implants, and imminent cancellous bone failure was implied when AP thickness was reduced to 6mm. Conclusion: UKRs with all-polyethylene tibial components are associated with greater proximal tibial strains than metal backed implants and this is exacerbated in thinner implants. The clinical consequences of this are uncertain. Medial UKR implantation does alter proximal tibial GSRb, though this is not uniform and is independent of implant type. When GSRb increases it is associated with ongoing pain.

Since the development of knockout and transgenic mouse models, mice have become the most widely used mammalian animal model to study bone. Despite the advances in knowledge of bone biology and function that have occurred from use of mouse models, many studies use primarily qualitative techniques, which may result in overlooking important subtle pathophysiologic changes. The hypothesis of this dissertation is that quantitative techniques to measure bone structure and function could identify the physiologic role of carbonic anhydrase II and calbindin-D28k in mouse bone, despite earlier qualitative studies indicating mice without these proteins have normal bone structure and function. Furthermore, a method to quantify bone function in vivo will be tested in a mouse model.Although carbonic anhydrase II deficient mice are less severely affected than patients, the mice demonstrate features of osteopetrosis including metaphyseal widening and a 50% increase in trabecular bone volume. The mice partially compensate for inhibited osteoclast function by increasing osteoclast number.Calbindin-D28k knockout mice demonstrated an increase in bone volume that results from additional bone at the endosteal surfaces. The higher bone volume results in increased stiffness and failure loads, highlighting the potential use of drugs that inhibit calbindin-D28k to treat diseases such as osteoporosis.Finally, calcium phosphate ceramic and hydroxyapatite particles used as strain gauge coatings demonstrated bone bonding to mouse femora after two months in vivo. The use of hydroxyapatite particles to coat strain gauges is the first time this method has been used with all commercially available materials, and will allow other research groups to use this technique. The major limitation to in vivo bone strain measurement in mice is the relatively large size of the sensors, which resulted in increased second moments of inertia in the implanted bones.Overall, this dissertation demonstrates that the use of quantitative techniques, including histology, histomorphometry, µCT imaging, and mechanical testing can measure subtle changes in bone properties that have been previously overlooked. Development of additional quantitative methods to study bone biomechanics in mouse models may encourage other research groups to quantify bone properties if no changes are noted using primarily qualitative methods.

No description available.