Engineering spinal fusion: evaluating ceramic materials for cell based tissue engineered approaches

Abstract

The principal aim of this thesis was to advance the development of tissue engineered posterolateral spinal fusion by investigating the potential of calcium phosphate ceramic materials to support cell based tissue engineered bone formation. This was accomplished by developing several novel model systems in a large animal (dutch milk goat) capable of simultaneously examining multiple materials in ectopic and orthotopic environments relevant to spinal fusion. Four calcium phosphate materials, hydroxyapatite (HA, at low and high sintering temperatures), biphasic calcium phosphate (BCP) and beta-tricalcium phosphate (TCP), were applied consistently throughout these models. In addition, the useof fluorochorme labels in the models enabled the dynamics of the initial bone formation processes to be observed and measured.

A purely ectopic model in the paraspinal muscle compared standardized scaffolds tissue engineered with bone marrow stromal cells (BMSCs). Standardization, achieved using an indirect rapid prototyping technique, was done to reduce variations typically seen in porous ceramics manufactured by traditional means in order to normalize the influence that these variations may have on cell behavior. Additionally, a screening model using cages implanted on decorticated transverse processes was developed specifically to engage both the bone and muscular aspects involved when generating a bony fusion between transverse processes in posterolateral spinal fusion. Two variants of this screening model examined the influence of material condition alone under idealized circumstances and the influence of tissue engineering in more clinically relevant porous materials. The idealized model created five uniform conduction channels per cage by separating two plates of each material by a small distance. Implanting two cages per animal in ten animals enabled ten different material conditions to be compared using a randomized complete block statistical design. For the second variation, the cages were adapted to enable four material blocks to be partitioned within each cage in order to compare the influence of porous ceramic materials on BMSC tissue engineering. Both cage models considered bone formation close to the underlying bone of the transverse process and adjacent to the overlying muscular tissue.

All models, ectopic scaffold and cages, demonstrated that high sintering temperature HA had the lowest bioactivity and was least affected by the addition of BMSCs. BCP and TCP were the most bioactive and most influenced by the addition of BMSCs (with the exception of TCP in the ectopic scaffold model). Furthermore, the cage model with porous blocks showed no influence of added BMSCs in areas close to the underlying bone (more orthotopic) but a significant beneficial effect in areas furthest from the underlying bone (more ectopic). This suggests that the addition of BMSCs may be beneficial to posterolateral spinal fusion where a majority of non-unions occur in the intertransverse space between the processes. The models we have developed and presented in this thesis represent reliable and efficient tools for screening multiple conditions in environments relevant to spinal fusion.