Page 133 - Visions of the Future Chemistry and Life Science

122 M. C. H. VAN DER MEULEN AND P. J. PRENDERGAST

signals the osteoblasts and osteoclasts either to add or to resorb tissue to
regain the physiological environment the cell requires. To maintain
normal bone mass, the sensing cells require a desired or reference stimu-
lus value. If the actual stimulus present in the tissue is less than the refer-
ence level, bone mass will be lost through resorption by osteoclasts, and if
the actual stimulus is above the reference level, bone will be formed by
osteoblasts. As a result of this adaptive response, the stimulus in the tissue
will approach and ultimately equal the desired stimulus value. Since the
sensory cells are distributed throughout the tissue, this model describes a
spatially discrete process in which each cell regulates its mechanical
stimuli by changing the mass or density of its extracellular environment.
The driving mechanical stimulus is not known, and many biomechanical
measures have been proposed, including strain, strain energy density, and
fatigue microdamage. These approaches can be coupled to computational
stress analysis procedures and have been used to predict bone adaptation
around implants and simulate the influence of mechanics on long bone
growth.
Recently, considerable interest has been centered on investigations of
the nonlinear dynamics of bone adaptation. Finite element models have
been used in iterative computer simulation procedures based on the feed-
back approach described above. The evolution of bone density and struc-
ture can be simulated for a given mechanical stimulus and initial
density pattern (Figure 7.6). This phenomenon can be viewed as a self-
organisational process operating within the bone: all elements will either
become fully dense or resorb to zero density, creating a porous ‘trabecular’
structure. The evolution of this density depends on the initial conditions,
so that a different initial density leads to a different trabecular structure,
indicating a nonlinear process. Furthermore, the final configuration is
metastable because a perturbation of the structure caused by the fracture
of a trabecula, for example, will not be followed by a return to the former
equilibrium. In reality, however, bone structures are stable because the
inevitable trabecular microfractures that occur in aged osteoporotic bone
do not lead to immediate degeneration, but rather the regulatory process
leads to a completely new equilibrium. If this computer simulation does
indeed capture the essence of bone adaptation, then adaptation is a far-
from-equilibrium dynamic process generated by positive feedback. To date,
these approaches have focused attention on the central role of mechanical
factors in determining bone structure.