Vous êtes sur la page 1sur 14

Journal of Biomechanics 35 (2002) 401–414

ESB Keynote LectureFDublin 2000


Why mechanobiology?
A survey article
Marjolein C.H. van der Meulena,b, Rik Huiskesc,*
a
Sibley School of Mechanical & Aerospace Engineering, Cornell University, Ithaca, NY 14853, USA
b
Department of Biomedical Mechanics and Materials, The Hospital for Special Surgery, New York, NY 10021, USA
c
Department of Biomedical Engineering, Eindhoven University of Technology, Wh 4.131, PO Box 513, 5600 MB Eindhoven, Netherlands
Accepted 5 September 2001

Abstract

The central paradigm of skeletal mechanobiology is that mechanical forces modulate morphological and structural fitness of the
skeletal tissuesFbone, cartilage, ligament and tendon. Traditionally, skeletal biomechanics has focussed on how these tissues
perform the structural and locomotory functions of the vertebrate skeleton. In mechanobiology the central question is how these
same load-bearing tissues are produced, maintained and adapted by cells as an active response to biophysical stimuli in their
environment. The idea that ‘form follows function’ is not new, but we now believe that the scientific community has the knowledge
and tools to prove, understand and use functional adaptation to benefit medicine and human health. In this Survey Article the
philosophy and progress of skeletal mechanobiology are discussed. The revival of this science, with roots dating back to the 19th
Century, is now driven by new developments in cellular, molecular and computational technologies. These developments are still in
an early stage of application, but if modern mechanobiology fulfills the promises of its ambitions, the results will bring great benefits
to tissue engineering and to the treatment and prevention of skeletal conditions such as congenital deformities, osteoporosis,
osteoarthritis and bone fractures. r 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Tissue differentiation; Bone adaptation; Computer simulation; Tissue engineering; Bone diseases

1. Introduction alization, chondrocyte apoptosis and bone modeling


proceeds spatially from the mid-diaphysis towards
Skeletal growth and development is an intricately the epiphyses. As the primary center progresses axially,
choreographed process involving many cell and tissue in some bones a secondary center forms, which
types (Fig. 1). During growth in the embryo, the long mineralizes through the same endochondral process.
bones initially form as mesenchymal condensations The growth plate is where the two growth fronts meet.
which chondrify to become cartilaginous anlagen. As Longitudinal growth at the growth plate occurs through
development progresses, the cartilage cellsFchondro- chondrocyte division and hypertrophy, and continues
cytesFhypertrophy and their extracellular matrix until skeletal maturity. Concurrently, radial diaphyseal
mineralizes, forming the primary ossification center. growth occurs through direct apposition of bone by
The chondrocytes die by apoptosis and the mineralized osteoblasts on the periosteal surface and resorption by
cartilage is modeled to bone by the coordinated activity osteoclasts on the endosteal surface. In parallel with the
of bone forming cells, osteoblasts, and bone resorbing formation of the mineralized skeleton, layers of cartilage
cells, osteoclasts. Osteoblasts that continue to produce at the joint surfaces develop into articular cartilage and
extracellular matrix and become trapped within the fibrous anlagen develop into tendons and ligaments.
matrix develop into mature bone cells, osteocytes. The Joints form at the interzones between the cartilage
cycle of chondrocyte hypertrophy, extracellular miner- anlagen.
The premise of mechanobiology is that these biolo-
*Corresponding author. Tel.: +31-40-247-4060; fax: +31-40-244-
gical processes are regulated by signals to cells generated
7355. by mechanical loading, a concept dating back to Roux
E-mail address: r.huiskes@tue.nl (R. Huiskes). (1881). The relevant questions include how external and

0021-9290/02/$ - see front matter r 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 2 1 - 9 2 9 0 ( 0 1 ) 0 0 1 8 4 - 1
402 M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414

Development of bone

A B C D E F
Embryonic Postnatal

Fig. 1. Schematic illustration of the development of long bones from embryonic to mature (figures not to scale). (A) Cartilaginous anlage with
chondrocytes; (B) Chondrocytes in the center swell; (C) Cartilage mineralization occurs around hypertrophic chondrocytes, proliferating flattened
cells develop; (D) Blood vessels penetrate the tissue, mineralized tissue is resorbed, bone formation takes place, and longitudinal growth commences;
(E) Secondary ossification centers develop and growth plates remain in between; (F) The bone is mature, the growth plates are closed; courtesy of Dr.
E. Tanck (2001).

muscle loads are transferred to the tissues, how the cells an optimized structure of minimal weight and maximal
sense these loads, and how the signals are translated into strength (Wolff, 1892).
the cascade of biochemical reactions to produce cell After the tissue has matured, bone is continually
expression or differentiation. Ultimately, we want to maintained by osteoclastic resorption and subsequent
predict growth and differentiation in quantitative terms, osteoblastic formation in a process called remodeling
based on a given force exerted on a given tissue matrix (Frost, 1987). One purpose of this process is the removal
populated by cells. of microcracks and microdamage (Verborgt et al., 2000),
While understanding the mechanobiology of bone hence the term ‘maintenance’. Osteoclasts are believed
development and growth can be relevant for the to be attracted by microcracks, through the apoptosis of
prevention and treatment of congenital deformities, osteocytes (Noble et al., 1997). At maturity, bone
there is also another motive for its study. The processes modeling continues due to variations in physical
of bone development continue in the mature organism activity; exercise enhances bone mass, while inactivity
during bone regeneration in fracture healing and during reduces it. In the elderly osteoporosis can develop, the
skeletal adaptation to implants. These involve tissue etiology of which is also thought to be affected by
differentiation from granulation through fibrous con- mechanobiology. And finally, the development of
nective tissue and cartilage to bone. The differentiation osteoarthritis in the elderly may be related to mechanical
pathway from granulation tissue to cartilage is a gradual loading as well.
process that can be influenced over time. Mineralization, To understand how a mechanical stimulus produces a
on the other hand, is a ‘catastrophic event’, occurring biological signal for cells to differentiate or adapt a
almost instantly and irreversibly. This event transforms tissue, we need to understand the relevant stimuli, the
the mechanical and morphological status of the tissue signal transduction pathways and the response processes
dramatically and completely, not a single chondrocyte (Duncan and Turner, 1995). Here, we described some
remains afterwards. After ossification, bone differentia- experimental and analytical approaches to address these
tion continues within the tissue. Osteoclasts and questions, with examples from work examining skeletal
osteoblasts now build the characteristic microstructures adaptation and differentiation.
of cortical or cancellous bone tissue. This process of As in many sciences, the integration of experimental
structural modulation and growth is called modeling and analytical models is critical to gaining an under-
(Frost, 1987). Mechanics also modulates the bone standing of the skeletal response to mechanical factors.
modeling process, as evident from the trabecular Experiments provide insights and measurements, which
patterns that line up with external forces, suggesting can then be interpreted within the context of analytical
M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414 403

frameworks. Analytical simulations permit investigation subjected to cyclic loading after initial healing demon-
of possible explanations that require in vivo validation strate significantly greater increases in stiffness and
and will suggest further experimental investigations. produce larger calluses (Goodship and Kenwright,
These investigations are greatly impacted by recent 1985). Cyclic compression produces stronger, but more
technological advances in imaging, computational me- compliant, healing fractures than static compression
chanics, genetics and molecular biology. The integration (Panjabi et al., 1979). Excessive strains inhibit the
of these techniques will provide many important insights normal healing process, producing less bone tissue and
into skeletal development and diseases. more initial inflammatory tissue (Augat et al., 1998).
However, growth factors and other chemical stimuli
likely play a more critical regulatory role than mechan-
2. Experimental mechanobiology ical stimuli, confounding these purely mechanobiologi-
cal models. Understanding the interaction between
Adaptation experiments examining skeletal mechan- growth factor expression and mechanical stimuli
obiology have been performed for more than a century may provide significant insights into the healing
and have examined the influence of both increasing and process. Computer models and simulations can aid our
decreasing the habitual loads applied to the skeleton. understanding of these fracture experiments (Blenman
Yet, despite the multitude of studies that have been !
et al., 1989; Bailon-Plaza and van der Meulen, 2001;
completed, there are still many unanswered questions Lacroix et al., 2000). Distraction osteogenesis is a
regarding the skeletal response to normal, perturbed and similar process to fracture healing and can also be used
pathological mechanical loading. Some of the difficulties to study tissue differentiation (Carter et al., 1998; Tay
arise from the design of experiments, others in the et al., 1998).
interpretation of data, and many from the inability to In vivo experiments have been used to examine both
compare across different species, ages, bones, loading tissue differentiation and adaptation to altered loading,
protocols, and so forth. Many of these issues arise from with the majority of experiments examining tissue
the inherent challenge of in vivo experiments. An remodeling in response to mechanical stimuli. Skeletal
introduction to in vivo bone mechanobiology experi- tissue differentiation experiments are complex to design;
ments, what is known from these experiments and the however, historically many ‘natural’ examples have been
open questions are presented below. Emphasis is placed cited (Pauwels, 1941; Wolff, 1892). The most common
on bone adaptation models due to the greater presence experimental design is a device that isolates or controls
in the literature. However, analogous questions can be the mechanical stimuli and then examines the tissues
posed for skeletal tissue differentiation and the adapta- that form (Goodman et al., 1994; Guldberg et al., 1997a;
tion of other musculoskeletal tissues in response to Sballe et al., 1992; Tagil and Aspenberg, 1999). When
mechanical loading. Underlying this discussion is the empty chambers are implanted into skeletal sites, new
fundamental goal of understanding skeletal mechano- tissue forms often in a process similar to fracture
transduction, which can be considered to consist of a healing. Bone can be formed both through intramem-
mechanical signal, response pathway and cellular branous formation (Guldberg et al., 1997a) and
adaptive response. endochondral ossification (de Rooij et al., 2001). Often
fibrous tissue forms within the chambers under parti-
2.1. Skeletal tissue differentiation cular loading protocols (Goodman et al., 1994). Once
the process for a given model is established, cellular and
Fracture healing is a natural skeletal tissue differ- tissue differentiation can be examined in the absence of
entiation process. A well-defined temporal-spatial tissue loading or with the application of controlled micromo-
generation and replacement pathway is present normally tion or loads in the device. These experiments demon-
during secondary fracture healing. Three phases follow strated the sensitivity of differentiating tissues to the
bone injury: an initial inflammatory phase, a tissue mechanical stimuli in their environment. With loading,
formation phase and finally tissue remodeling. Biophy- different tissues form (de Rooij et al., 2001) and the
sical stimuli are integral to the healing process. This can morphology of the tissues also differs (Guldberg et al.,
already be appreciated by comparing primary fracture 1997a). While the intent is a simple loading situation,
healing to secondary healing. Primary healing is this is difficult to realize in these in vivo devices due to
completely stable and forms bone directly (Einhorn, complex boundary conditions, tissue interfaces and
1998). Secondary healing involves lower fragment nonlinear behavior of the tissues, in addition to the
stability and larger gap sizes and heals through continually evolving nature of these conditions. Again,
successive tissue formation and replacement (Einhorn, mechanical-hypothesis-based theories can help explain
1998). Experiments have demonstrated a significant role and interpret these tissue differentiation models, as
of mechanical forces in the process of secondary fracture discussed below (Yuan et al., 2000; Prendergast et al.,
healing. When compared to rigid fixation, fractures 1997; Giori et al., 1995)
404 M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414

2.2. Bone functional adaptation studies (Li et al., 1991). Whether these differences were
caused by differences in the local mechanical environ-
Bone adaptation to mechanical stimuli has been ment or by fundamentally different adaptive responses is
studied extensively using in vivo models. This subject unknown. Answering these questions experimentally is
is a subset of the broader examination of cell and tissue difficult because the in vivo loads cannot be measured
differentiation and may be a more straightforward directly. In adult animals, physiological exercise gen-
problem as a starting point for our understanding, erally produced bone hypertrophy or no response. The
because only a single tissue is under consideration, bone. absence of a response may indicate that the exercise
Bone adaptation has been studied using a wide variety protocol did not sufficiently elevate the loads above
of experimental models. Approaches used to increase habitual levels. These unknowns all point towards the
loading include exercise, osteotomy and devices to apply need for a better understanding of the applied mechan-
controlled loading (Biewener and Bertram, 1993; ical loads.
Churches and Howlett, 1982). Decreased loading was To better address the mechanical stimulus underlying
achieved by casting, neurectomy, hindlimb suspension bone adaptation, noninvasive experimental models have
and space flight (Morey-Holten et al., 1996; Uhthoff and been developed recently, which control the stimulus
Jaworksi, 1978; Vico et al., 2001). Several common applied to the cortical diaphysis (Gross et al., 2000;
concepts emerge from the past century of in vivo Mosley et al., 1997; Torrance et al., 1994; Turner et al.,
experimentation. We know that bone adaptation occurs 1991). The two most common, currently, are the ulnar
in response to cyclic, not static, loading. The outcome is compression model of Lanyon and coworkers (Fig. 2)
mainly changes in cortical bone quantity, not bone and the tibial four-point bending approach developed at
quality. Cancellous adaptation is less well characterized Creighton by Turner and colleagues. The ulnar com-
but has similar responses that are present as changes in pression model likely creates more physiological stresses
the tissue apparent density. In general, demonstrating a and strains within the bone and has load application
definitive increase in bone mass or strength is more points that are farther from the point of interest than the
difficult than demonstrating a loss. Increased properties tibial four-point bending model. Both models were
are also harder to achieve with physiological than initially developed for the rat and have been modified
nonphysiological loading conditions. Finally, we know for the mouse (Akhter et al., 1998; Fritton et al., 2001).
that the adaptive response is generally greater in Neither model requires surgery as previous controlled
growing than mature animals, but growth may con- loading models often did (Churches and Howlett, 1982;
found the results. Hert et al., 1969; Rubin and Lanyon, 1984). In both
While these qualitative observations generally hold systems, dynamic loads are applied and the load
true, precise quantitative predictions of the skeletal magnitude, rate, number of cycles and duration are
response to mechanical stimuli are often desirable and well controlled. These noninvasive models, combined
currently impossible. Much of our inability to predict with increased computing power, higher resolution
adaptation arises from experimental logistics: until
recently, most skeletal functional adaptation experi-
ments could not identify the driving feature of the
mechanical stimulus nor the contributing elements of
the response pathway (Duncan and Turner, 1995). For
example, when the loading environment is altered by
exercise and a skeletal effect is observed, does the
response result from the increased applied forces, the
increased numbers of loading cycles, the increased
loading rates, another exercise-induced parameter, or a
combination of these and other factors? Can one design
an exercise study to distinguish these features of the
stimulus? Increasing bone mass and strength with in
vivo physiological loading such as exercise is notoriously
difficult. Exercise studies have produced contradictory
results: running chickens at one speed was shown to
increase cortical area and moments of inertia (Biewener
et al., 1986), while running for a longer duration at a
similar intensity suppressed bone growth and reduced
strength over a similar experimental period (Matsuda
et al., 1986). Differences from one skeletal site to the Fig. 2. The ulnar compression model (from Mosley and Lanyon,
next within the same animal were also found in these 1998).
M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414 405

imaging and new molecular and genetic techniques conditions; simply characterizing the normal phenotype
enable systematic evaluation of loading parameters to may be unrevealing or insufficient. However, manipulat-
understand the nature of the osteogenic stimuli and ing the system by changing the applied loads can require
pathways. For the cortical diaphysis, stimulus charac- cellular/tissue adaptation above and beyond normal
teristics that induce bone modeling are becoming development and can demonstrate significant phenoty-
evident. The applied strain rates and magnitudes need pic effects. Examples include the osteopontin-knockout
to be high (Mosley et al., 1997; Mosley and Lanyon, mouse and the bone morphogenetic protein-5 deficient
1998), the strain rate contributes to the morphology of short ear mouse. The former has no obvious phenotype,
the bone formed (Turner et al., 1994), distinct, but does not lose bone when subjected to hindlimb
temporally separated loading ‘episodes’ are required, suspension (Ishijima et al., 2001). The phenotype of the
not just increased numbers of loading cycles (Robling latter is an overall reduced body size and these mice
et al., 2000), and gender and genetic background respond to increased loads more slowly than hetero-
influence the nature of the response (Akhter et al., zygous controls (Mikic et al., 1996; van der Meulen,
1998; Pedersen et al., 1999). 1999). Unfortunately, developing transgenic or knock-
Similar questions need to be examined in cancellous out mice and examining all the potential regulatory
bone, a site of more clinical relevance than the cortical factors will consume many research careers. However,
diaphyses. Applying controlled loading to cancellous further strengths of these genetic models will become
bone is more difficult than to the cortical diaphyses. evident when we start to examine genetic control of
Trabecular bone adaptation models include direct skeletal responses by not only expression but also
stimulation of the site of interest as well as in linkage analyses (Beamer et al., 2001; Klein et al.,
encapsulating bone chambers (Chambers et al., 1993; 2001; Yershov et al., 2001).
Goldstein et al., 1991, 1997b; Guldberg et al., 1997b; In summary, many questions remain regarding the
Hollister et al., 1996; Lamerigts et al., 2000; Morgan role of mechanics in the formation and adaptation of
et al., 2001). Nearly all models require surgical inter- skeletal tissues. For functional adaptation, we need to
vention at the trabecular site; the rat-tail vertebra model be able to globally describe the adaptive response for
is an exception, since only the adjoining vertebral bodies different skeletal sites and ages, for example. To do so
are operated on. Systematic application of dynamic will require understanding the nature of an osteogenic
loading needs to be performed to address the same mechanical stimulus and characterizing the molecular
questions posed above for cortical bone. For example, pathway that this stimulus activates. More generally, in
the rodent ulna model can presumably be extended also skeletal tissue formation these same questions apply,
to examine cancellous bone, and perhaps also the focused on the mechanical stimuli that induce cell and
response of the cartilage in the growth plate to tissue differentiation. Answering these questions will
mechanical loading. require carefully controlled experiments with well-
Compared to questions about the nature of the characterized mechanical environments. When design-
stimulus, even less is known regarding the transduction ing and interpreting experiments, consideration must be
of osteogenic signals to stimulate bone formation or given to contributing factors, including the type of
resorption. This pathway is hard to delineate, particu- model (in vitro or in vivo), species, type of loading
larly when using in vivo models. However, recent (physiological or invasive), the nature of loading
molecular advances allow examination of whether (tension/compression, cyclic/static, controlled/natural,
individual genes and molecules contribute to the adapta- etc.) and level of analysis (cell, organ or tissue level).
tion pathways. The mouse provides a unique opportunity The discussion here has focused on in vivo models with
to examine genetic factors through manipulation of an emphasis on controlled applied loading. A similar
murine embryonic stem cells and because significant discussion could be presented for in vitro experiments,
portions of the genome have been characterized (Clark which are used to answer similar questions, and subject
and Rowe, 1996). Transgenic technology has led to the to different limitations.
creation of both knockout and transgenic animals, either
missing or over-expressing genes, and previously re-
corded natural mutations can now be identified. One can 3. Computational mechanobiology
now directly determine whether a given protein is
important to skeletal adaptation by removing it. The purpose of computational mechanobiology is to
Transgenic technology allows the role of potential determine the quantitative rules that govern the effects
regulatory factors (Erlebacher and Derynck, 1996), cell of mechanical loading on tissue differentiation, growth,
surface receptors (Zimmerman et al., 2000) and matrix adaptation and maintenance by trial-and-error. From a
constituents (Khillan et al., 1991) to be examined. mechanics point of view, the task is considered a
Increasingly, mouse genetic studies are demonstrating ‘boundary value problem’, whereby the boundary loads
extreme redundancy within the genome under normal of a domain are translated into local mechanical
406 M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414

variables within the domain, depending on the geometry dilatational and distortional strain components. The
and mechanical properties of its materials. Such first invariant represents a change in volume (as caused
problems are usually solved with finite element analysis by dilatational or hydrostatic stress), the second a
(FEA). The biological side of the computation is based change in shape (as caused by distortional or shear
on the premise that local mechanical variables stimulate stress). Pauwels proposed that the combination of these
cell expression to regulate matrix composition, density variables determines the differentiation pathway from
or structure. These two parts of the problem are the mesenchymal origin to fibrous, fibro-cartilaginous or
combined in a computer simulation model. Modeling cartilaginous tissue. Endochondral bone formation
considerations include force application at the bound- would be stimulated in areas of low distortional and
ary, force transmission through the tissue matrix, high dilatational strain (Prendergast and van der
mechanosensation by cells, mechanotransduction by Meulen, 2001).1 Pauwels’ theory was based on clinical
the cells, cell gene expression, and transformation of observation and logic, but he did not have the capacity
extracellular matrix characteristics. These processes to measure or calculate the tissue strains or stresses
must be represented in variables, parameters and accurately. In the following discussion we will show how
mathematical relationships. Some of these are known, the application and evolution of FEA has affected our
or can be measured (e.g. morphology, mechanical tissue thoughts about Pauwels’ ideas, and how computational
properties, external loading characteristics), but others mechanobiology is gradually maturing.
cannot. So how are these unknowns dealt with in the Carter and colleagues tested Pauwels’ theory relative
computations? to endochondral bone formation using FEA. Endo-
There is a difference between the output of a chondral ossification was defined by the peak cyclic
computation and the output of a problem to be solved. hydrostatic stress (D) and the peak cyclic octahedral
For example, the local mechanical variables that might shear stress (S), combined in an ‘osteogenic index’ I, as
stimulate the cells are usually not known and neither are
I ¼ S þ kD;
the mathematical relationships between a given stimulus
value and its effect on local cell expression and matrix summed over a number of loading cycles; k is an
adaptation. Potential stimuli include strain, fluid flow, empirical constant. High values of the index were
strain rate and strain energy. Although these unknowns hypothesized to stimulate ossification; no specific
must be entered into the computer simulation, they also threshold value was specified. Note that S is always
represent precisely the quantitative rules that we are positive, but D can both be positive (tension) or negative
trying to determine. Computational mechanobiologists (compression). This tissue differentiation rule was tested
hypothesize a potential rule and determine if the with an FEA model of a developing joint to determine
outcome of this hypothesis produces realistic tissue whether this approach could predict the emergence of
structures and morphologies, hence ‘trial-and-error’. If secondary ossification centers (Carter and Wong, 1988).
the results correspond well, they might be an explana- High values of the osteogenic index were found in the
tion for the mechanism being modeled. This method of cartilage precisely where osteogenesis occurs during in
research is common practiceFand productiveFin vivo skeletal development. Comparing the results with
physics, less common in biology (Huiskes, 1995); biological examples, the most predictive values for the
although ‘theoretical biology’ is based on this type of constant k were determined at 0:3pkp1:0 (Fig. 3).
approach. Computer simulation has recently been cited Similarly successful predictions of endochondral ossifi-
as the ‘third method of science’, after logic and cation sites were performed for other skeletal sites, for
experimentation (Kelly, 1998). Below we give some bone regeneration in fracture healing and for healing
examples from skeletal development, repair, mainte- around orthopaedic implants (Carter et al., 1998; Carter
nance and adaptation, particularly to demonstrate the and Beaupre! , 2001; Giori et al., 1995).
general approach, its limitations and its progress in A vital issue for mechanobiological FEA models is to
recent years. what extent they can be simplified without loosing their
potential to obtain meaningful results. On the one hand,
3.1. Tissue differentiation simplicity is dictated by contemporary computational
technology, and on the other it depends on the
Tissue differentiation occurs in bone development and
regeneration, fracture healing and adaptation to skeletal 1
The inclusion of enchondral ossification in a theoretical framework
implants. In 1941, Pauwels proposed a theoretical for skeletal differentiation pathways is not entirely justified, as
framework in which the effects of mechanical forces mineralization and bone modeling are very different kinds of processes,
which actually terminate differentiation. Mineralized tissue cannot de-
on tissue differentiation pathways occur through me-
differentiate, only resorb; or rather be resorbed. However, mineraliza-
chanical deformation of the tissues. The local deforma- tion occurs when particular conditions in cartilage are met and the
tions in a tissue can be described by a strain tensor, creation of those conditions can be seen as a part of the differentiation
which has first and second invariants representing the pathway.
M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414 407

analyses of bone ossification suggested that the true fluid


pressure in the tissue might be sensitive to the solid
matrix compressibility, another poorly understood
parameter (Tanck et al., 1999). Based on a comparative
analysis of experimental tissue mineralization studies
(Burger et al., 1991) with biphasic FEA, the hypothesis
that hydrostatic stress was a modulating variable was
refuted (Tanck et al., 2000). In a biphasic analysis of a
tissue differentiation experiment around a moving
piston in the femoral condyles of dogs (Sballe et al.,
1992), Prendergast and Huiskes (1996), Prendergast and
colleagues (1997) found that local tissue fluid pressure
does not change as the tissue differentiates. Yuan et al.
(2000) came to a similar conclusion from a biphasic
analysis of soft-tissue membranes under tibial knee
implants, comparing results to those of an earlier linear-
elastic analysis of the same problem (Giori et al., 1995).
Prendergast and Huiskes (1996), Prendergast et al.
Fig. 3. The distributions of the osteogenic index as computed in the (1997) also found that the stresses on cells are not only
cartilaginous part of developing bone with a 2-D FEA model, using 3 generated by the tissue matrix, but also to a large extent
different values for the empirical constant k (Carter and Wong, 1988).
by the drag forces from interstitial fluid flow. This
Ossification is predicted where the index is high (from Carter, 1987).
conclusion indicates the need for dynamically loaded,
biphasic models, because these effects cannot be
examined with static or linear elastic ones. Interstitial
hypothesis to be tested. In that sense a model takes for fluid flow and pressure need to be investigated as
granted the matter in dispute, or ‘begs the question’, potential signaling variables in the simulations.
because its assumptions determine its answers. One can Improvements in computer capacity now enable FEA
only make conclusions about phenomena accounted for computer simulations in a dynamic sense, as opposed to
in the model, not about what is omitted or assumed not the earlier computer analyses. In the latter case the
to contribute. This is equally true for experimental mechanical variables are computed for a particular
models of course. The above FE analyses of the cartilage initial configuration and predict a steady state, final
in the models assumed homogeneity, isotropy and linear homeostatic configuration. The result is determined
elasticity; while in reality the tissue is multi-phasic, from the initial conditions, but the process is not
nonlinear and highly structured at many levels of scale. simulated over time. The results of Carter and associates
Hydrostatic pressure and octahedral stress in biphasic (Carter and Beaupre! , 2001) relative to the prediction of
tissues seemed to be well captured by single-phasic linear ossification areas are quite realistic. However, these
elastic FEA, provided that the permeability is low, as is analyses cannot predict the progression of ossification
the case in cartilage (Mow et al., 1980; Tanck et al., based on a single set of initial conditions. To simulate
2000; Yuan et al., 2000). Hence, relative to the the time-varying mineralization process, the evolution of
hypothesis the authors wanted to test, the conclusions the tissue material properties must also be simulated
about the predictive quality of the osteogenic index may with time. The process of ossification itself affects the
not have been any different had they described the load transfer in the cartilaginous parts while time
cartilage as biphasic. But, although the proposed transpires (Claes and Heigele, 1999; Gardner et al.,
definition of the osteogenic index produces realistic 2000; Lerner et al., 1998; Prendergast et al., 1997;
results, another choice for other mechanical variables, Stevens et al., 1999). Another advantage of time-
or an alternative mathematical rule, might produce dependent simulations is the ability to incorporate
equal or potentially even better results. dynamic loading, including both fluctuating loads at a
One important consideration is fluid flow in cartilage. particular instant and overall variation in load over
There are several reasons that interstitial fluid flow time.
could be a more realistic mechanical variable for Based on the above considerations, and the studies of
feedback information to the cells during tissue differ- Prendergast and Huiskes (1996), Prendergast et al.
entiation than hydrostatic pressure. Flow is known to (1997), an alternative rule was proposed for differentia-
stimulate anabolic cell expressions during in vitro testing tion from loose connective tissue to cartilage or bone.
(Jacobs et al., 1998). The hydrostatic stress, computed in This was tested in dynamic, biphasic computer simula-
linear-elastic FEA models, is believed to be indicative of tions as a mechano-regulation index, deciding which
the interstitial fluid pressure in vivo. However, biphasic tissue phenotype would prevail as an effect of mechan-
408 M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414

ical stimulation. For that purpose a mechano-regulation because the piston displacement was force controlled
index (M) was formulated as (Huiskes et al., 1997) (Prendergast et al., 1997). Force control reduces the
g n excursion of the piston when the tissue gets stiffer,
M¼ þ ; allowing the tissue to control its differentiation stimulus.
a b
To test this conclusion, simulations of motion control
where g ðdimensionlessÞ is the maximal distortional (or were performed using the maximal total excursion as in
octahedral shear) strain, v (mm/s) the interstitial fluid- the force-controlled experiment, regardless of the force
flow velocity, a ¼ 0:0375 and b ¼ 3 (mm/s). Based on required. While the tissue ossified fully in the force-
experimental data (Sballe et al., 1992), it was proposed controlled simulation, motion control produced a
that where M > 3; fibrous tissue would prevail (char- largely fibro-cartilaginous tissue, with only a few islands
acterized by a modulus of 2.0 MPa and a permeability of of bone (Fig. 5). This lack of ingrowth under large
1.0  1014 m4 N/ s), where 1pMp3; cartilage would displacements is also seen around loose orthopedic
prevail (modulus 10.0, permeability 5.0  1015) and implants.
where Mo1; bone would form (modulus 4590, perme- These results demonstrate the importance of temporal
ability 3.7  1013). This theory has two empirical simulation for the questions posed. In both the force-
thresholds which determine when transitions in the controlled and motion-controlled simulations of tissue
tissue type occur, and two empirical constants a and b ingrowth the differentiation patterns predicted after the
which are, however, interdependent. In this model the first iterations were identical, because both the force and
value b=a plays a similar role as k in the osteogenic excursion of the piston were equal (Fig. 5). Therefore, in
index. The rule was tested in a mechano-regulatory an analysis based on the first iteration only, the same
computer simulation, using the FEA model of the outcome would have been predicted for both cases. The
Sballe experiment discussed above (Huiskes et al., same regulatory model was used to simulate tissue
1997). Iteratively, cyclic loading was applied to the differentiation in fracture healing, using the same values
piston and the tissue modulus and permeability were for the empirical constants and tissue characteristics
adjusted locally per iteration, in accordance with the (Lacroix et al., 2000). In these simulations the differ-
distribution of the mechano-regulation index in the gap entiation pathways were also not monotonic, again
tissue, until differentiation converged to a final tissue demonstrating the importance of simulation over
phenotype (Fig. 4). Using this approach, the differentia- analyses based on the initial conditions only.
tion of gap-tissue phenotype from fibrous connective to While this simulation model captures the dynamic
cartilage to bone could be predicted, simulating the aspects of the system and accounts for the time-
experiment (Sballe et al., 1992). In the earlier analysis dependent properties of the tissues, limitations are
of this experiment the ingrowth process was successful present as well. The fluid pressure and flow velocity
can be evaluated, but osmotic effects and charged-
density flows in the tissue, for example, are not
New Tissue Phenotype accounted for (Mow et al., 1999). Hence, this model
Finite Element Model Mechanical Characteristics
Implant Loading Conditions
also ‘begs the question’. Similar to the linear elastic,
homeostatic models, the relevant mechanical variables
Mechano-Regulation Diagram are evaluated at a macroscopic (homogenized) con-
tinuum level and cannot describe what individual cells
Fluid Velocity

FIBROUS experience. Cells exist at a much lower scale, surrounded


by a quite irregular micro-morphology. These models
GAP
FIBRO are ‘mechanistic’ with respect to fluid flow, but very
TON

CARTILAGE
crude and approximate with regard to the cellular
PIS

BONE
environment (Wang et al., 2001). But, as stated earlier,
BONE from a scientific point of view, these models could still
Strain produce realistic hypotheses, because the importance of
these details is unknown. The physiological cellular
mechanisms are not understood, so the appropriateness
Tissue Strain
Fluid-Solid Velocity
of macroscopic or microscopic formulations is un-
known. Perhaps the extracellular matrix acts as a filter
Fig. 4. The regulatory scheme used in the computer simulation. The for the local environment, making a continuum model
strain and fluid-velocity distributions are calculated in the FEA model. equally appropriate. To capture every detail of the
Based on the transition criteria, illustrated in a phase diagram
system from cells to whole organs, we may need to build
representing the differentiation rule, tissue phenotype characteristics
are updated in every iteration. The simulation is continued until no FEA models of whole bones with refinement to the
more change occurs and homeostasis is reached (from Huiskes et al., cellular level, including the cascade of biochemical
1997). reactions in and around the cells, but current computer
M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414 409

3
10

fluid-solid velocity [um/s]


2
10

1
10
Element status after first iteration
0
10 of both force-controlled and motion-
-1
controlled piston actuation
10 -4 -3 -2 -1
10 10 10 10
strain [-]

Final after force-control Final after motion-control


3 3
10 10

fluid-solid velocity [um/s]


fluid-solid velocity [um/s]

2 2
10 Fibrous tissue 10

1 1
10 10
Fibro-cartilage
0 0
10 10
Bone
-1
-1 10 -4
10 -4 -3 -2 -1 10
-3
10 10
-2 -1
10
10 10 10 10
strain [-] strain [-]
Fig. 5. Graphs illustrating the phenotypical stages of the gap tissue during the simulation; every dot represents an element. Initially all elements are
fibrous. After the first iteration two elements were predicted to become bone, a large number to become fibro-cartilaginous and the majority to be still
fibrous, based on the strain and fluid velocity history computed. This first phase developed identical in both the force and motion controlled
simulations. In the final homeostatic stage of the force-controlled simulation all elements turned to bone (Huiskes et al., 1997). In the motion-
controlled simulation, however, the gap tissue was predicted as a final mixture of fibrous and fibro-cartilaginous, with just a few islands of bone.

capacity still defies that. There is nothing scientifically values of these constants can only be determined by
incorrect about the current approaches, however. This comparing computational results to similar experiments.
discussion relates to the extent a model is ‘mechanistic’, If one assumes that the cell mediated processes are
i.e. to what detail is the biological cascade of processes, independent of cell location, then the empirical con-
and their biochemical actors, represented or necessary to stants should have unique values for all differentiation
be represented? In computational mechanobiology processes. If the empirical constants must be adjusted
mechanistic detail is not a goal by itself; the research per configuration or problem definition, then a theory is
question dictates the level of sophistication required. not falsifiable, and has little predictive power (Prender-
A final complicating issue for computational models gast, 2001). Comparing results of several different FEA
in mechanobiology is the presence of empirical con- models to evaluate the ossification index to bone-
stants, whose values must be determined by comparison formation patterns in ontogenesis, growth-plate orienta-
to a biological reality. The osteogenic ossification rule tion, fracture healing, and fibrous-tissue characteristics
(Carter and Beaupre! , 2001), for example, contains the around orthopaedic implants, the ossification rule
constant k, which weights the sensitivity of the tissue for delivered best-fitting k-values between 0.0 and 2.0
hydrostatic stress relative to shear stress in the index. In (Carter and Beaupre! , 2001; Ribble et al., 2001). As can
the mechano-regulation rule (Huiskes et al., 1997; be seen in Fig. 3, the value of k can have a large effect on
Lacroix et al., 2000) the constant b/a weights the relative the predicted distribution of the osteogenic index. The
sensitivities for flow velocity and strain. In addition, differentiation rule was only tested for ingrowth of
both theories require transitional values for their implants, as shown above, and fracture healing (Lacroix
indexes, which determine when the tissue differentiates; et al., 2000), using the same values for the empirical
the ossification rule requires a single value for this constants. Much more experimentation will be needed
purpose and the differentiation rule requires two. The to validate these values.
410 M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414

Fig. 6. A model to simulate trabecular modeling, remodeling and adaptation governed by strain-energy density rate as produced by external loads,
assuming osteocyte mechanosensation and signal transduction to lining cells at the trabecular surface, which stimulates osteoblastic bone formation.
Osteoclasts are assumed to remove microcracks in a spatially random way; resorption cavities produce the local strain concentrations attracting
osteoblastic repair. An example is shown in which, after 2000 iterations with the FEA model, a homeostatic structure, adapted to magnitude and
orientation of the external load, was formed from an arbitrary initial structure (from Huiskes et al., 2000).

Tissue differentiation encompasses a wide area of differentiation. Hart (2001) recently wrote an excellent
scientific efforts, including many computational studies historical overview of theories and computational
that were not discussed here; relevant ones can be found models. Modern developments started with the ‘theory
in the articular cartilage area (Wang et al., 2001). The of adaptive elasticity’ for cortical bone (Cowin and
testing of theories for tissue differentiation with simula- Hegedus, 1976), which was later used for an FEA
tion models has spread to ligaments and tendons (Wren simulation model by Hart et al. (1984). These studies
et al., 1998) and development of articular joints examined the adaptation of cortical form to external
(Heegaard et al., 1999). Currently the theories proposed forces. Alternative theories were proposed and used by
can only be used to predict the past, allowing the Huiskes et al. (1987), Mattheck et al. (1998), Prender-
formulations to be validated against known experiments gast and Taylor (1994) and van der Meulen et al. (1993).
or biological realities. Our ultimate objective should be Adaptation of (trabecular) bone density was simulated
to predict the future. by several groups, using different isotropic (Beaupre!
et al., 1990; Carter et al., 1987; Huiskes et al., 1987;
3.2. Bone adaptation Weinans et al., 1992a, b) or anisotropic theories (Hart
and Fritton, 1997; Jacobs et al., 1997; Luo et al., 1995).
Computer-simulation of bone adaptation is more All these trabecular simulation models were phenomen-
mature than in the tissue differentiation area, not in the ological in nature and simulated the outcome of
least because the problems are better defined and coordinated osteoclastic and osteoblastic activity as
concern only bone, as mentioned earlier. For this case either a net increase or decrease in density. A more
the computational studies also started by analysis, with mechanistic approach is necessary to represent the true
true simulation following more recently. One could say physiology of the adaptation process, which occurs only
that Wolff’s Law (1892) was already based on the on trabecular surfaces (Huiskes, 2000). A model to
comparison between trabecular morphology and the simulate local trabecular adaptation that does allow this
results of a ‘computational’ analysis (Huiskes, 2000). surface mechanism to be simulated, and also includes a
The philosophy behind the computer simulation studies biological osteocyte mechanosensory and signaling
in this area is similar to those discussed above for tissue function, was developed by Mullender and Huiskes
M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414 411

(1995). A model that also includes separate descriptions for interpreting experiments and experiments for in-
of osteoclastic resorption and osteoblastic formation, sights and observation to develop models; yet another
enabling simulation of mechanobiologically modulated circular mechanism! The challenge is to educate
growth, adaptation and maintenance (remodeling), scientists who can understand and contribute to both
illustrated in Fig. 6, was published recently (Huiskes ends of the spectrum and help to eliminate the
et al., 2000; Ruimerman et al., 2001). Adachi et al. distinction. Many other fields rely upon this integration.
(2001) proposed a phenomenological adaptation model In mechanobiology, however, the interdisciplinary
for the trabecular structure using nonuniformity of the nature of these questions cuts across many additional
local surface stress distribution as the regulatory feed- spectra besides experiments and simulations: in vivo to
back signal. in vitro, biology to engineering, laboratory to clinic, etc.
The models incorporating trabecular morphology This not only produces great challenges but also
have been used only to study adaptation in small cubes potentially significant rewards.
of bone. Analyses of whole bone FEA models with Beyond the integration within the field itself, we must
refinement down to the trabecular level are feasible now ask where mechanobiology fits in a broader scientific
(van Rietbergen et al., 1999). However, there is not and social context. Scientifically, mechanobiology is
enough computer capacity yet to use those in iterative central to unraveling ‘nature versus nurture’ distinc-
simulations to study trabecular adaptation based on a tions. With the sequencing of the human genome
(mechanistic) cell modulation scheme. Future models complete (Venter et al., 2001) it is now apparent that
will need to include both natural boundary conditions at the genetic code is only the beginning and provides few
external bone surfaces and refined trabecular architec- answers to how form is generated and maintained. The
ture. limitations of genomics and proteomics emphasize the
importance of understanding the role of environmental
influences, particularly biophysical factors (Stewart,
4. Discussion 1998). Scientific interest in form and function is not
new, but recent advances in cell and molecular biology
So why mechanobiology? While the term is the inverse allow examination of phenomena that were previously
of biomechanics, the definition of biomechanics clearly inaccessible.
encompasses the experimental and computational mod- On Earth, gravity is ubiquitous and is likely one of the
els described here: ‘The science that studies the effects of most pervasive factors in the development of tissues.
forces on biological tissues, organs and organisms, in This particularly holds true for the musculoskeletal
relation to biological and medical problems’’ (ESB, system, which plays a central role in vertebrate’s ability
1978). As a word, however, mechanobiology moves the to function in Earth’s gravitational environment. The
emphasis from mechanics to biology and to determining potential reward for successful quantitative models and
the mechanisms behind ‘form follows function’. But insightful experiments is enormous. The challenge lies in
then what does ‘function’ follow? In an evolutionary disentangling environmental modulation and genetic
sense, one could consider function to follow form, a predisposition in the skeleton. The conservation of
strategy that has enabled vertebrates to successfully osseous microstructure and morphology speak, to a high
breed and survive. This then produces ‘yform degree, of predisposition. Yet, the action of gravity and
follows function follows formy’ Considering this the resulting mechanical stimuli are also highly con-
cycle of repetition, one can now distinguish between served, making a case for modulation. Successfully
the two terms: biomechanics focuses on the latter part, distinguishing adaptation from genetic programming of
whether function follows form, whereas mechanobiol- cell metabolism will be a great mechanobiological feat,
ogy emphasizes the former, whether function determines requiring intimate interaction between experimental and
form. computational efforts.
Both experimental and computational studies are The potential for mechanobiology to contribute to
critical to advancing our understanding of form and clinical progress is also promising. Mechanically based
function, and even more important is the integration of diseases such as osteoporosis and osteoarthritis are
models and experiments. Yet we have separated experi- already areas of intense research activity. In addition,
mental and computational studies in this article, a the development of successful synthetic and engineered
distinction that reflects contemporary science and organs and tissues, ‘tissue engineering’, will depend on
presumably arises from the cultural roots of both mechanobiological progress. This is especially true for
activities. Experimental studies, particularly in vivo functional tissue engineering (Butler et al., 2000). Not
models, originated in biology, while computational only is an intimate knowledge of the natural tissues
work is fundamentally based on physics and engineer- necessary, but controlling and modulating mechanical
ing. To make significant progress, the objective needs to stimuli will be essential to develop appropriately
be the integration of these two types of studies: models engineered organs and to integrate and function with
412 M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414

the host. The realization of this promise will be Carter, D.R., Fyhrie, D.P., Whalen, R.T., 1987. Trabecular bone
particularly rewarding. density and loading history: regulation of connective tissue biology
by mechanical energy. Journal of Biomechanics 20, 785–794.
Carter, D.R., Beaupr!e, G.S., Giori, N.J., Helms, J.A., 1998.
Mechanobiology of skeletal regeneration. Clinical Orthopaedics
Acknowledgements 355, S41–S55.
Chambers, T.J., Evans, M., Gardner, T.N., Turner-Smith, A., Chow,
This article synthesizes two presentations from the J.W.M., 1993. Induction of bone formation in rat tail vertebrae by
Mechanobiology Symposium of the 12th Conference of mechanical loading. Bone Minerals 20, 167–178.
Churches, A.E., Howlett, C.R., 1982. Functional adaptation
the European Society of Biomechanics in Dublin,
of bone in response to sinusoidally varying controlled compressive
Ireland, August 28–30, 2000. We thank the ESB and loading of the ovine metacarpus. Clinical Orthopaedics 168, 265–
the organizing committee for the opportunity to present 280.
our work. Funding is acknowledged from the National Claes, L.E., Heigele, C.A., 1999. Magnitudes of local stress and strain
Institutes of Health (USA) and the Netherlands along bony surfaces predict the course and type of fracture healing.
Foundation for Research (NWO-MW). Journal of Biomechanics 32, 255–266.
Clark, S., Rowe, D.W., 1996. Transgenic animals. In: Bilizikian, J.P.,
Rais, L.G., Rodan, G.A. (Eds.), Principles of Bone Biology.
Academic Press, New York, pp. 1161–1172.
Cowin, S.C., Hegedus, D.H., 1976. Bone remodeling I: theory of
References
adaptive elasticity. Journal of Elasticity 6, 313–326.
de Rooij, P.P., Siebrecht, M.A.N., T.agil, M., Aspenberg, P., 2001. The
Adachi, T., Tsubota, K., Tomita, Y., Hollister, S.J., 2001. Trabecular fate of mechanically induced cartilage in an unloaded environment.
surface remodeling simulation for cancellous bone using micro- Journal of Biomechanics 34, 961–966.
structural voxel finite element models. Journal of Biomechanical Duncan, R.L., Turner, C.H., 1995. Mechanotransduction and the
Engineering 123, 403–409. functional response of bone to mechanical strain. Calcified Tissue
Akhter, M.P., Cullen, D.M, Pedersen, E.A., Kimmel, D.B., Recker,
International 57, 344–358.
R.R., 1998. Bone response to in vivo mechanical loading in two
Einhorn, T.A., 1998. The cell and molecular biology of fracture
breeds of mice. Calcified Tissue International 63, 442–449.
healing. Clinical Orthopaedics 355, S7–S21.
Augat, P., Margevicius, K., Simon, J., Wolf, S., Suger, G., Claes, L.,
Erlebacher, A., Derynck, R., 1996. Increased expression of TGF-beta 2
1998. Local tissue properties in bone healing: influence of size and
in osteoblasts results in an osteoporosis-like phenotype. Journal of
stability of the osteotomy gap. Journal of Orthopaedic Research
Cell Biol 132, 195–210.
16, 475–481.
ESB, 1978. Inaugural document for the formation of the European
!
Bailon-Plaza, A., van der Meulen, M.C.H., 2001. A mathematical
Society of Biomechanics. ESB Archives.
framework to study the effects of growth factor influences on
Frost, H.M., 1987. Vital biomechanics. Proposed general concepts for
fracture healing. Journal of Theortical Biology 212, 191–209.
skeletal adaptation to mechanical usage. Calcified Tissue Interna-
Beamer, W.G., Schultz, K.L., Donahue, L.R., Churchill, G.A., Sen, S.,
tional 45, 145–156.
Wergedal, J.R., Baylink, D.J., Rosen, C.J., 2001. Quantitative trait
Fritton, J.C., Myers, E.R., van der Meulen, M.C.H., Bostrom,
loci for femoral and lumbar vertebral bone mineral density in
M.P.G., Wright, T.M., 2001. Validation of a loading apparatus:
C57BL/6J and C3H/HeJ inbred strains of mice. Journal of Bone
characterization of murine tibial surface strains in vivo. Transac-
and Mineral Research 16, 1195–1206.
Beaupr!e, G.S., Orr, T.E., Carter, D.R., 1990. An approach for time- tions of the Orthopaedic Research Society 26, 535.
dependent bone modeling and remodelingFtheoretical develop- Gardner, T.N., Stoll, T., Marks, L., Mishra, S., Knothe, T.M., 2000.
ment. Journal of Orthopaedic Research 8, 651–661. The influence of mechanical stimulus on the pattern of tissue
Biewener, A.A., Bertram, J.E.A., 1993. Mechanical loading and bone differentiation in a long bone fractureFan FEM study. Journal of
growth in vivo. In: Hall B.K. (Ed.), Bone, Vol. 7. Boca Raton FL, Biomechanics 33, 415–425.
CRC Press, pp. 1–36. Giori, N.J., Ryd, L., Carter, D.R., 1995. Mechanical influences on
Biewener, A.A., Swartz, S.M., Bertram, J.E.A., 1986. Bone modeling tissue differentiation at bone-cement interfaces. Journal of Ar-
during growth: dynamic strain equilibrium in the chick tibiotarsus. throplasty 10, 514–522.
Calcified Tissue International 39, 390–395. Goldstein, S.A., Matthews, L.S., Kuhn, J.L., Hollister, S.J., 1991.
Blenman, P.R., Carter, D.R., Beaupre, G.S., 1989. Role of mechanical Trabecular bone remodeling: an experimental model. Journal of
loading in the progressive ossification of a fracture callus. Journal Biomechanics 24 (Suppl 1), 135–150.
of Orthopaedic Research 7, 398–407. Goodman, S.B., Song, Y., Doshi, A., Aspenberg, P., 1994. Cessation
Burger, E.H., Klein-Nulend, J., Veldhuijzen, J.P., 1991. Modulation of of strain facilitates bone formation in the micromotion chamber
osteogenesis in fetal bone rudiments by mechanical stress in vitro. implanted in the rabbit tibia. Biomaterials 15, 889–893.
Journal of Biomechanics 24 (Suppl 1), 101–109. Goodship, A.E., Kenwright, J., 1985. The influence of induced
Butler, D.L., Goldstein, S.A., Guilak, F., 2000. Functional tissue micromovement upon the healing of tibial fractures. Journal of
engineering: the role of biomechanics. Journal of Biomechanical Bone and Joint Surgery 67B, 650–655.
Engineering 122, 570–575. Gross, T., Srinivasan, S., Bailey, M., Bain, S., 2000. Noninvasive
Carter, D.R., 1987. Mechanical loading history and skeletal biology. activation of bone formation in the murine tibia. Transactions of
Journal of Biomechanics 20, 1095–1109. the Orthopaedic Research Society 25, 149.
Carter, D.R., Beaupr!e, G.S., 2001. Skeletal Function and Form: Guldberg, R.E., Caldwell, N.J., Guo, X.W., Goulet, R.W., Hollister,
Mechanobiology of Skeletal Development, Aging and Regenera- S.J., Goldstein, S.A., 1997a. Mechanical stimulation of tissue
tion. Cambridge University Press, Cambridge. repair in the hydraulic bone chamber. Journal of Bone and Mineral
Carter, D.R., Wong, M., 1988. The role of mechanical loading Research 12, 1295.
histories in the development of diarthrodial joints. Journal of Guldberg, R.E., Richards, M., Caldwell, N.J., Kuelske, C.L., Gold-
Orthopaedic Research 6, 804–816. stein, S.A., 1997b. Trabecular bone adaptation to variations in
M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414 413

porous-coated implant technology. Journal of Biomechanics 30, Lacroix, D., Prendergast, P.J., Marsh, D., Li, G., 2000. A model to
147–153. simulate the regenerative and resorption phases of fracture healing.
Hart, R.T., 2001. Bone modeling and remodeling: theories and Transactions of the Orthopaedic Research Society 25, 869.
computation. In: Cowin, S.C. (Ed.), Bone Biomechanics Hand- Lerner, A.L., Kuhn, J.L., Hollister, S.J., 1998. Are regional variations
book. CRC-Press, Boca Raton, pp. 1–42 (Chapter 31). in bone growth related to mechanical stress and strain parameters?
Hart, R.T., Fritton, S.P., 1997. Introduction to finite element based Journal of Biomechanics 31, 327–869335.
simulation of functional adaptation of cancellous bone. Forma 12, Li, K.-C., Zernicke, R.F., Barnard, R.J., Li, A.F.-Y., 1991. Differential
277–299. response of rat limb bones to strenuous exercise. Journal of
Hert, J., Liskov!a, M., Landrgot, B., 1969. Influence of the long- Applied Physiology 70, 554–560.
term, continuous bending on the bone. An experimental study on Luo, G., Cowin, S.C., Sadegh, A.M., Arramon, Y.P., 1995.
the tibia of the rabbit. Folia Morphologica (Prague) 17, Implementation of strain rate as a bone remodeling stimulus.
389–399. Journal of Biomechanical Engineering 117, 329–338.
Hart, R.T., Davy, D.T., Heiple, K.G., 1984. A computational method Mattheck, C., 1998. Design in Nature. Springer, Berlin.
for stress analysis of adaptive elastic materials with a view toward Matsuda, J.J., Zernicke, R.F., Vailas, A.C., Pedrini, V.A., Pedrini-
applications in strain-induced bone remodeling. Journal of Mille, A., Maynard, J.A., 1986. Structural and mechanical
Biomechanical Engineering 106, 342–350. adaptation of immature bone to strenuous exercise. Journal of
Heegaard, J.H., Beaupr!e, G.S., Carter, D.R., 1999. Mechanically Applied Physiology 60, 2028–2034.
modulated cartilage growth may regulate joint surface morphogen- Mikic, B., van der Meulen, M.C.H., Kingsley, D.M., Carter, D.R.,
esis. Journal of Orthopaedics Research 17, 509–517. 1996. Mechanical and geometric changes in the growing femora of
Hollister, S.J., Guldberg, R.E., Kuleske, C.L., Caldwell, N.J., BMP-5 deficient mice. Bone 18, 601–608.
Richards, M., Goldstein, S.A., 1996. Relative effects of wound Morey-Holton, E.R., Whalen, R.T., Arnaud, S.B., van der Meulen,
healing and mechanical stimulus on early bone response to porous- M.C., 1996. The skeleton and its adaptation to gravity. In: Blatteis,
coated implants. Journal of Orthopaedic Research 14, 654–662. C.M., Fregly, M.J. (Eds.), Handbook of Physiology: Environ-
Huiskes, R., 1995. The law of adaptive bone remodeling: a case for mental Physiology, Part III: The Gravitational Environment,
crying newton? In: Odgaard, A., Weinans, &H. (Eds.), Bone Section 1: Microgravity. Oxford University Press, Oxford, pp.
Structure and Remodeling. World Scientific Publishing, Singapore, 691–719.
River Edge, London, pp. 15–24. Morgan, T.G., Yang, X., Baldini, T., van der Meulen, M.C.H., Myers,
Huiskes, R., 2000. If bone is the answer, then what is the question? E.R., Wright, T.M., Bostrom, M.P.G., 2001. Trabecular bone
Journal of Anatomy 197, 145–156. adaptation to controlled in vivo loading. Transactions of the
Huiskes, R., Weinans, H., Grootenboer, H.J., Dalstra, M., Fudala, B., Orthopaedics Research Society 26, 238.
Slooff, T.J., 1987. Adaptive bone-remodelling theory applied to Mosley, J.R., Lanyon, L.E., 1998. Strain rate as a controlling influence
prosthetic-design analysis. Journal of Biomechanics 20, 1135–1151. on adaptive modeling in response to dynamic loading of the unla in
Huiskes, R., van Driel, W.D., Prendergast, P.J., Sballe, K., 1997. A growing male rats. Bone 23, 313–318.
biomechanical model for periprosthetic fibrous-tissue differentia- Mosley, J.R., March, B.M., Lynch, J., Lanyon, L.E., 1997. Strain
tion. Journal of Material Science: Materials in Medicine 8, 785– magnitude related changes in whole bone architecture in growing
788. rats. Bone 20, 191–238198.
Huiskes, R., Ruimerman, R., van Lenthe, G.H., Janssen, J.D., 2000. Mow, V.C., Kuei, S.C., Lai, W.M., Amstrong, C.G., 1980. Biphasic
Effects of mechanical forces on maintenance and adaptation of creep and stress relaxation of articular cartilage in compression:
form in trabecular bone. Nature 405, 704–706. theory and experiment. Journal of Biomechanical Engineering 102,
Ishijima, M., Rittling, S.R., Yamashita, T., Tsuji, K., Kurosawa, H., 73–84.
Nifuji, A., Denhardt, D.T., Noda, M., 2001. Enhancement of Mow, V.C., Wang, C.C.-B., Hung, C.T., 1999. The extracellular
osteoclastic bone resorption and suppression of osteoblastic bone matrix, interstitial fluid and ions as a mechanical signal transducer
formation in response to reduced mechanical stress do not occur in in articular cartilage. Osteoarthritis & Cartilage 7, 41–58.
the absence of osteopontin. Journal of Experimental Medicine 193, Mullender, M.G., Huiskes, R., 1995. A proposal for the regulatory
399–404. mechanism of Wolff’s law. Journal of Orthopaedic Research 13,
Jacobs, C.R., Simo, J.C., Beaupr!e, G.S., Carter, D.R., 1997. Adaptive 503–512.
bone remodeling incorporating simultaneous density and aniso- Noble, B.S., Stevens, H., Reeve, J., Loveridge, N., 1997. Identification
tropy considerations. Journal of Biomechanics 30, 603–613. of apoptotic changes in osteocytes in normal and pathological
Jacobs, C.R., Yellowley, C.E., Davis, B.R., Zhou, Z., Cimbala, J.M., human bone. Bone 20, 273–282.
Donahue, H.J., 1998. Differential effect of steady versus oscillating Panjabi, M.M., White III, A.A., Wolf Jr., J.W., 1979. A biomechanical
flow on bone cells. Journal of Biomechanics 31, 969–976. comparison of the effects of constant and cyclic compression on
Kelly, J., 1998. The third culture. Science 279, 992–993. fracture healing in rabbit long bones. Acta Orthopaedica Scandi-
Khillan, J.S., Olsen, A.S., Kontusaari, S., Sokolov, B., Prockop, D.F., navica 50, 653–661.
1991. Transgenic mice that express a mini-gene version of the Pauwels, F., 1941. Grundri einer Biomechanik der Frakturheilung.
human gene for type I procollagen. (COL1A1) develop a In: 34th Kongress der Deutschen Orthop.adischen Gesellschaft.
phenotype resembling a lethal form of osteogenesis imperfecta. Ferdinand Enke Verlag, Stuttgart. (Biomechanics of the Locomo-
Journal of Biological Chemistry 266, 23373–23379. tor Apparatus, translated by Manquet, P., Furlong, R. (Eds.),
Klein, R.F., Shea, M., Gunness, M.E., Pelz, G.B., Belknap, J.K., 1980, Springer, Berlin, pp. 375–407).
Orwoll, E.S., 2001. Phenotypic characterization of mice bred for Pedersen, E.A., Akhter, M., Cullen, D.M., Kimmel, D.B., Recker,
high and low peak bone mass. Journal of Bone and Mineral R.R., 1999. Bone response to in vivo mechanical loading in C3H/
Research 16, 63–71. HeJ mice. Calcified Tissue International 65, 41–46.
Lamerigts, N.M.P., Buma, P., Huiskes, R., Schreurs, W., Gardeniers, Prendergast, P.J., 2001. An analysis of theories in biomechanics.
J., Slooff, T.J.J.H., 2000. Incorporation of morsellized bone graft Engineering Transactions 49, 117–133.
under controlled loading conditions. A new animal model in the Prendergast, P.J., Huiskes, R., 1996. Finite element analysis of fibrous
goat. Biomaterials 21, 741–747. tissue morphogenesisFa study of the osteogenic index using a
414 M.C.H. van der Meulen, R. Huiskes / Journal of Biomechanics 35 (2002) 401–414

biphasic approach. Mechanics of Composite Materials 32, 209– modeling response uncomplicated by trauma or periosteal pressure.
218. Calcified Tissue International 54, 241–247.
Prendergast, P.J., Taylor, D., 1994. Prediction of bone adaptation Turner, C.H., Akhter, M.P., Raab, D.M., Kimmel, D.B., Recker,
using damage accumulation. Journal of Biomechanics 27, 1067– R.R., 1991. A noninvasive, in vivo model for studying strain
1076. adaptive bone modeling. Bone 12, 73–79.
Prendergast, P.J., Huiskes, R., Sballe, K., 1997. Biophysical stimuli Turner, C.H., Forwood, M.R., Rho, J.Y., Yoshikawa, T., 1994.
on cells during tissue differentiation at implant interfaces. Journal Mechanical loading thresholds for lamellar and woven bone
of Biomechanics 30, 539–548. formation. Journal of Bone and Mineral Research 9, 87–97.
Prendergast, P.J., van der Meulen, C.H., 2001. Mechanics of bone Uhthoff, H.K., Jaworski, Z.F.G., 1978. Bone loss in response to long-
regeneration. In: Cowin, S.C. (Ed.), Bone Biomechanics Hand- term immobilisation. Journal of Bone and Joint Surgery 60B, 420–
book. CRC-Press, Boca Raton, pp. 1–13 (Chapter 32). 429.
Ribble, Th.G., Santare, M.H., Miller, F., 2001. Stresses in the growth van der Meulen, M.C.H., 1999. Differential bone adaptation to
plate of the developing proximal femur. Journal of Applied mechanical loading: role of bone morphogenetic protein-5. 1999
Biomechanics 17, 129–141. Bioengineering Conference. ASME, New York, pp. 689–690.
Robling, A.G., Burr, D.B., Turner, C.H., 2000. Partitioning a daily van der Meulen, M.C.H., Beaupr!e, G.S., Carter, D.R., 1993.
mechanical stimulus into discrete loading bouts improves the Mechanobiologic influences in long bone cross-sectional growth.
osteogenic response to loading. Journal of Bone and Mineral Bone 14, 635–642.
Research 15, 1596–1602. .
van Rietbergen, B., Muller, .
R., Ulrich, D., Ruegsegger, P., Huiskes,
Roux, W., 1881. Der Kampf der Teile im Organismus. Engelmann, R., 1999. Tissue stress and strain in trabeculae of a canine proximal
Leipzig. femur can be quantified from computer reconstructions. Journal of
Rubin, C.T., Lanyon, L.E., 1984. Regulation of bone formation by Biomechanics 32, 443–451.
applied dynamic loads. Journal of Bone and Joint Surgery 66A, Venter, J.C., Adams, M.D., Myers, E.W., et al., 2001. The sequence of
397–402. the human genome. Science 291, 1304–1351.
Ruimerman, R., Huiskes, R., van Lenthe, H., Janssen, J.D., 2001. A Verborgt, O., Gibson, G.J., Schaffler, M.B., 2000. Loss of osteocyte
computer-simulation model relating bone-cell metabolism to integrity in association with microdamage and bone remodeling
mechanical adaptation of trabecular architecture. Computer after fatigue in vivo. Journal of Bone and Mineral Research 15, 60–
Methods in Biomedical and Bioengineering 4, 433–448. 70.
.
Sballe, K., Hansen, E.S., B-Rasmussen, H., Jrgensen, P.H., Bunger, Vico, L., Hinsenkamp, M., Jones, D., Marie, P.P.J., Zallone, A.,
C., 1992. Tissue ingrowth into titanium and hydroxyapatite coated Cancedda, R., 2001. Osteobiology, strain, and microgravity. Part II
implants during stable and unstable mechanical conditions. Journal Studies at the tissue level. Calcified Tissue International 68, 1–10.
of Orthopaedic Research 10, 285–299. Wang, C.C.B., Hung, C.T., Mow, V.C., 2001. An analysis of the effects
Stevens, S.S., Beaupr!e, G.S., Carter, D.R., 1999. Computer model of of depth-dependent aggregate modulus on articular cartilage stress-
endochondral growth and ossification in long bones: biological and relaxation behavior in compression. Journal of Biomechanics 34,
mechanobiological influences. Journal of Orthopaedic Research 17, 75–84.
646–653. Weinans, H., Huiskes, R., Grootenboer, H.J., 1992a. The behavior of
Stewart, I., 1998. Life’s Other Secret. Allen Lane/The Penguin Press, adaptive bone–remodeling simulation models. Journal of Biome-
London. chanics 25, 1425–1441.
T.agil, M., Aspenberg, P., 1999. Cartilage induction by controlled Weinans, H., Huiskes, R., Grootenboer, H.J., 1992b. Effects of
mechanical stimulation in vivo. Journal of Orthopaedic Research material properties of femoral hip components on bone remodel-
17, 200–204. ing. Journal of Orthopaedics Research 10, 845–853.
Tanck, E. 2001. Mechanical regulation of bone development. Ph.D. Wolff, J., 1892. The Law of Bone Remodeling. (Das Gesetz der
Thesis, University of Nijmegen, The Netherlands. Transformation der Knochen, Kirschwald, 1892). Translated by
Tanck, E., van Driel, W.D., Hagen, J.W., Burger, E.H., Blankevoort, Maquet, P., Furlong, R. Springer, Berlin, 1986.
L., Huiskes, R., 1999. Why does intermittent hydrostatic pressure Wren, T.A., Beaupr!e, G.S., Carter, D.R., 1998. A model for loading-
enhance the mineralization process in fetal cartilage? Journal of dependent growth, development, and adaptation of tendons and
Biomechanics 32, 153–161. (Associated Letters-to-the-Editors, J ligaments. Journal of Biomechanics 31, 107–114.
Biomechanics 32, 1255–1257. Yershov, Y., Baldini, T.H., Villagomez, S., Young, T., Martin, M.L.,
Tanck, E., Blankevoort, L., Haaijman, A., Burger, E.H., Huiskes, R., Bockman, R.S., Peterson, M.G., Blank, R.D., 2001. Bone strength
2000. The influence of muscular activity on local mineralization and related traits in HcB/Dem recombinant congenic mice. Journal
patterns in metatarsals of the embryonic mouse. Journal of of Bone and Mineral Research 16, 992–1003.
Orthopaedic Research 18, 613–619. Yuan, X., Ryd, L., Huiskes, R., 2000. Wear particle diffusion and
Tay, B.K., Le, A.X., Gould, S.E., Helms, J.A., 1998. Histochemical tissue differentiation in TKA implant fibrous interfaces. Journal of
and molecular analyses of distraction osteogenesis in a mouse Biomechanics 33, 1279–1286.
model. Journal of Orthopaedic Research 16, 636–642. Zimmerman, D., Jin, F., Leboy, P., Hardy, S., Damsky, C., 2000.
Torrance, A.G., Mosley, J.R., Suswillo, R.F., Lanyon, L.E., 1994. Impaired bone formation in transgenic mice resulting from altered
Noninvasive loading of the rat ulna in vivo induces a strain-related integrin function in osteoblasts. Developmental Biology 220, 2–15.

Vous aimerez peut-être aussi