Most of what is known about fracture healing comes from studies of shaft fractures in long bones. In contrast, patients more often have fractures closer to the ends (metaphyses). Here most bone tissue has a spongy, cancellous structure different from the compact bone of the shaft. There is an increasing awareness that metaphyseal fractures heal differently. However, the more easily studied shaft healing has usually been considered as good enough representative for fracture healing in general.

My work shows that the biology of metaphyseal healing is more different from shaft healing than was previously known and that this has implications on the effect of various commonly prescribed drugs.

First we studied biopsies of healing cancellous bone collected from human donors. We found that the most abundant new bone formation occurred freely in the marrow rather than on the surface of old trabeculae, as described in most literature. There was little cartilage, indicating that the dominant bone formation process is mostly membranous in nature. This is a contrast to the ample cartilage formation commonly found in the well-characterized shaft fracture models.

Next we characterized a model that allows for mechanical quantification of regenerating cancellous bone. By contrasting this cancellous healing model with the standard shaft healing model we could demonstrate that the NSAID indomethacin, the glucocorticoid dexamethasone, and the bisphosphonate alendronate all had different effects on the mechanical quality of bone regeneration in shaft and metaphysis; while anti-inflammatory drugs strongly impaired shaft healing, metaphyseal healing was not similarly affected. Alendronate had a positive effect on both models, though the effect was strongest in the metaphyseal model. Taken together these differences shed some light as to the differences in healing biology.

The last step (within the boundaries of this thesis) was a characterization of how healing in cortical and cancellous bone differs in terms of immune cell involvement. We could find little difference between the two bone types day 3. However, day 5 an increase in the number of granulocytes could be noted in the cancellous bone while the cortical bone had a higher number of lymphocytes.

To conclude, this work furthers our understanding of how metaphyseal healing differs from shaft healing. It has clinical implications as it motivates an increased attention to the site of fracture while contemplating treatment. I hope this thesis can be read as an argument for increased interest in metaphyseal fracture healing.