Fixation - Orthopaedic Biomaterials in Research and Practice - page 329

resulting from remodeling. All cortical bone will turn over, although the
rate in adults is quite low, perhaps 3%-6% per year. However, adaptive
remodeling may increase this rate. In any case, the ability of these mate-
rials, especially the “bioactive” glasses, to form bonds to bone depends
on physical property changes that occur immediately after implanta-
tion. There have apparently been no studies done to determine whether
implant surfaces released by late remodeling are suitable substrates for
bonding of newly formed bone.
Bond strength
Bond strengths are difficult to compare because of the different struc-
tures of the interfaces. Cemented interfaces display shear strengths
between 0.1 MPa (non-precoated, failure at cement-implant interface)
and 10-12 MPa (well intruded, failure at cement-cancellous or cancel-
lous-cortical interface). Mature adhered interfaces appear to exceed
the strength of the adhered bone, producing an estimated strength
of 5-15 MPa (longitudinal shear, dependent on cortical bone density).
Ingrown interfaces produce strengths, which are somewhat dependent on
the surface coating structure and anatomic location, of 15-18 MPa (metal-
lic or ceramic coatings, 30% porosity, radial shear, failure in cortical
bone). Thus, the peak strengths seem to be limited by properties of the sup-
porting tissue, rather than by their intrinsic structural and compositional
differences. Furthermore, these apparent differences in strength of fully
mature interfaces may reflect difficulties in testing these different struc-
tures. Combined with the absence of any well-documented animal studies
comparing all three types contemporaneously, it is safest to assume that
all three possess the same ultimate strength, perhaps 10-15 MPa.
It is interesting to compare these values with the peak stresses shown
in Figure 13.5. The peak outer fiber tensile stresses, with an implant
in place, are calculated to be about 5 MPa (tensile lateral, compressive
medial). If it is assumed that the inner fiber cortical stresses are this
high (actually, they must be lower for geometric reasons) and the stem
is assumed to be infinitely stiff (worst case), then the interfacial shear
stress would not exceed about one-third of the single-cycle strength
of these interfaces. A more realistic estimate would probably lead to
the conclusion that mature, well-formed interfaces of any of the three
types are 10 times stronger than the normal peak load carriage requires.
Therefore, it is improbable that interfacial failure or “loosening” occurs
by a simple mechanical failure, although low-stress, high-cycle fatigue
failure remains a possibility.
“Loosening” or failure of fixation
If fixation does not fail by a primarily mechanical process, then its
origins are most probably biologic. That is, “loosening” represents an