Mediation of bone ingrowth in porous hydroxyapatite bone graft substitutes.

Previous investigations have shown that both the early biological response and the mechanical properties of a porous hydroxyapatite bone graft substitute are highly sensitive to its pore structure. The objective of this study was to evaluate whether the pore structure continued to influence bone integration in the medium to long term. Two screened batches of porous hydroxyapatite (PHA) designated as batch A and batch B, with porosities of approximately 60 and 80%, respectively, were selected for this study and implanted for periods of 5, 13, and 26 weeks into the lower femur of New Zealand White rabbits. Histomorphometric analysis of the absolute volume of bone ingrowth within batch A and B implants from 5 to 26 weeks showed that the absolute volume of bone ingrowth was consistently lower in batch A (10-21%), compared to batch B implants (24-31%). However, when the volume of bone ingrowth was normalised for the available pore space, this difference was reduced (23-47% and 32-42% for batches A and B, respectively). These observations suggest that differences in the volume of bone ingrowth initially depended on pore interconnectivity rather than pore size, whereas the volume or morphology of the PHA influenced the volume and morphology of bone ingrowth at later time points. Compression testing showed that bone ingrowth had a strong reinforcing effect on PHA bone graft substitutes, and a strong correlation was identified between mechanical properties and the absolute volume of ingrowth for both batches A and B. Furthermore, at 13 and 26 weeks, there was no significant variation in the ultimate compressive strength of integrated batch A and B implants. This similarity in ultimate mechanical properties indicated that the absolute volume of ingrowth may be mediated by the PHA structure through its impact on the dynamics of the local biomechanical environment. The results of push-out testing showed that fixation of PHA bone graft substitutes was independent of density within the range studied, with no significant difference in the interfacial shear stress between batches A and B at each time point throughout the study. Copyright 2003 Wiley Periodicals, Inc. J Biomed Mater Res 68A: 187-200, 2004