The overall goal of this research is to provide insight into the bone biomechanics of human lumbar vertebrae implanted with total disc arthroplasty (TDA) implants. The research elucidates fundamental mechanisms by which lumbar TDA implants alter the yield strength of the vertebral body, stress in the bone tissue, and trabecular-cortical load-sharing behavior. The findings could have broad clinical and scientific implications related to pre-operative assessment of bone quality, in vitro test protocols for device development and regulation, the development of new implant designs, and clinical exclusion criteria. Ultimately, the hope is that this insight can be used to improve clinical outcomes for this class of device, and, in turn, contribute to the broad effort of addressing the global health burden associated with degenerative spinal pathologies.

This dissertation comprises a series of computational experiments conducted on human cadaveric lumbar vertebrae using a combination of high-resolution micro-computed tomography image data with parallel linear and nonlinear finite element analysis implemented on a peta-scale supercomputing cluster. The goals of these studies are to assess mechanisms of load transfer and stress development in the bone, including how tissue-level stress depends on the loading mode of the implant, its size and material characteristics, and inter-individual variations in the structural features of the vertebral body.

The primary results indicate that the load-transfer behavior in the underlying vertebral bone is substantially altered by TDA, causing high levels of trabecular bone stress, diminishing the role of the cortical shell, and substantially reducing the whole-bone yield strength. These findings were assessed in relationship to the vertebral bone’s morphology and microstructure, enabling mechanistic insight into bone sub-failure and failure behavior. A key finding was that small declines in trabecular bone volume fraction can have a magnified effect on whole-bone yield strength following TDA;​ therefore, pre-operative assessment of bone quality should optimally focus on the trabecular centrum.

Another important finding was that when the footplate of a TDA implant is loaded through its anterior or posterior rim, representing load-transfer into the bone following impingement induced by flexion or extension, bone stress near the impinged region increases substantially. This localized stress increase causes a reduction in whole-bone yield strength, suggesting that subsidence might be caused, in some cases, by impingement-type loading. Pre-clinical subsidence testing of TDA implants should therefore include some form of evaluation of bending-induced impingement if the implant enables this type of loading in vivo. As a possible strategy to help reduce the risk of implant subsidence, future research might be directed at whether an implant could be designed to reduce the frequency or strength-reduction effects of impingement loading.

In closing, this dissertation elucidates fundamental mechanisms by which the vertebral bone tissue resists the loads applied by lumbar TDA implants. This mechanistic insight was used to identified possible targets—including pre-clinical tests and implant design considerations—that might help to improve clinical outcomes for this class of device