How can bone turnover modify bone strength independent of bone mass.pdf

Bone 42 (2008) 1014

–

1020

www.elsevier.com/locate/bone

Review

How can bone turnover modify bone strength independent of bone mass?

C.J. Hernandez ⁎

Musculoskeletal Mechanics and Materials Laboratory, Case Western Reserve University, Cleveland, OH, USA

Department of Mechanical and Aerospace Engineering, Case Western Reserve University, Cleveland, OH, USA

Department of Orthopaedics, Case Western Reserve University, Cleveland, OH, USA

Received 2 January 2008; revised 22 January 2008; accepted 2 February 2008

Available online 20 February 2008

Abstract

The amount of bone turnover in the skeleton has been identified as a predictor of fracture risk independent of areal bone mineral density (aBMD)

and is increasingly cited as an explanation for discrepancies between areal bone mineral density and fracture risk. A number of mechanisms have been

proposed to explain how bone turnover influences bone biomechanics, including regulation of tissue degree of mineralization, the disconnection or

fenestration of individual trabeculae by remodeling cavities, and the ability of cavities formed during the remodeling process to act as stress risers.

While these mechanisms can influence bone biomechanics, they also modify bone mass. If bone turnover is to explain any of the observed

discrepancies between fracture risk and areal bone mineral density, however, it must not only modify bone strength, but must also modify bone

strength in excess of what would be expected from the associated change in bone mass. This article summarizes biomechanical studies of how tissue

mineralization, trabecular disconnection, and the presence of remodeling cavities might have an effect on cancellous bone strength independent of

bone mass. Existing data support the idea that all of these factors may have a disproportionate effect on bone stiffness and/or strength, with the

exception of average tissue degree of mineralization, which may not affect bone strength independent of aBMD. Disproportionate effects of mineral

content on bone biomechanics may instead come from variation in tissue degree of mineralization at the micro-structural level. The biomechanical

explanation for the relationship between bone turnover and fracture incidence remains to be determined, but must be examined not in terms of bone

biomechanics, but in terms of bone biomechanics relative to bone mass.

Keywords: Bone turnover; Bone remodeling; Bone quality; Bone biomechanics

Contents

Introduction ............................................................... 1014

Bone volume ............................................................ 1015

Tissue degree of mineralization ................................................... 1016

Disconnection of trabeculae ..................................................... 1017

Remodeling cavities ......................................................... 1017

Conclusions ............................................................... 1018

Acknowledgments ............................................................ 1019

References ................................................................ 1019

Introduction

⁎ Glennan 615A, 10900 Euclid Avenue, Cleveland, OH 44106-7222, USA.

E-mail address: christopher.hernandez@case.edu .

Bone turnover represents the total volume of bone that is

both resorbed and formed over a period of time [1] . In adults,

bone turnover occurs primarily through bone remodeling, a

focal process that involves the coupled activity of osteoclasts

doi: 10.1016/j.bone.2008.02.001

C.J. Hernandez / Bone 42 (2008) 1014 – 1020

1015

Table 1

The reduction in human cancellous bone stiffness and strength associated with a 6% difference in bone volume or mass is shown

Process through which bone

remodeling modifies bone mass

Difference in

bone mass

Expected reduction

in stiffness

Expected reduction

in strength

Source

Reduction in bone volume

− 6% (volume)

12 – 16%

9 – 14%

Empirical power law models

[14,17 – 21]

Reduction in average tissue degree of

mineralization from 65% to 62%

ash by weight

− 6% (bone mineral content)

11 – 13%

95% confidence interval from

empirical power law models

[41]

Intraspecimen variation in tissue

degree of mineralization

0% reduction in bone mass;

increase in COV of tissue

mineralization from 20% to 50%

14 – 24%

Not yet evaluated Micro-computed tomography

based finite element models

[45,46]

− 6% (volume) 3 – 39% 18 – 35% 3D cellular solid finite element

models [50,51]

Addition of remodeling cavities − 6% (volume) 12 – 47% 13 – 61% Micro-computed tomography

based finite element models

[53,55]

The 95%confidence interval is reported if available, otherwise the range across all cited studies is shown. A factor that causes a greater change in bone stiffness or strength

than that caused by a 6% reduction in bone volume (the first row) has the potential to explain how bone turnover can influence fracture risk independent of bone quantity.

and osteoblasts [2] . Changes in the amount of bone turnover

cause local changes in bone volume and affect the average age of

tissue in a bone, resulting in alterations in tissue degree of

mineralization and trabecular microarchitecture. Clinical find-

ings that biochemical markers of bone turnover can predict

fracture risk independent of areal bone mineral density (aBMD)

have led to the suggestion that the amount of bone turnover in

the skeleton can have a biomechanical effect independent of

bone mass 1

summarized in Table 1 a 6% decline in bone mass was

selected as it is the estimated size of the remodeling space in

the spine [11] .

Bone volume

8] , and that bone turnover may help to explain

discrepancies between aBMD and fracture risk [5,6] . The

biomechanical effects of alterations in bone turnover are

commonly attributed to modifications in tissue degree of

mineralization, the fenestration or disconnection of individual

trabeculae, and/or remodeling cavities acting as stress risers

[4,6

–

Because the first step in the remodeling process is bone

resorption, each remodeling event is associated with the for-

mation of a temporary cavity. The total volume of bone oc-

cupied by all remodeling cavities and unmineralized bone tissue

(osteoid) is known collectively as the remodeling space [12] .

Increases in bone turnover result in an increase in the volume

occupied by the remodeling space and cause a corresponding

reduction in mineralized bone volume.

Biomechanical testing of cancellous bone specimens has

shown that bone stiffness and strength are related to apparent

density (

9] . While these mechanisms can influence bone strength,

they also modify bone mass. If one of these mechanisms is to

explain any of the discrepancies between aBMD and fracture

risk, however, it must have an effect on bone strength that is

much larger than would be expected from the change in bone

mass alone. The purpose of this article is to review the

biomechanical effects of changes in bone that can be caused

by bone turnover. The current article concentrates on the

biomechanics of human cancellous bone specimens 3

, g/cm 3 ) through power law relationships. Because

apparent density is directly related to bone volume fraction (BV/

TV) [13] , the same power law relationships are valid for bone

volume fraction as well. These power law relationships can be

expressed as follows:

5mmin

smallest dimension, as that is the scale at which the mecha-

nisms mentioned above all have biomechanical significance.

Biomechanics of bone at this size scale is also important

because a factor that does not have a disproportionate bio-

mechanical effect at this scale could not have a dispropor-

tionate biomechanical effect at the scale of the whole bone

[10] . The biomechanical effects of a 6% difference in bone

mass caused by different aspects of bone turnover are

–

~ q

;

r Ult ~ q

;

σ Ult is

the strength (ultimate stress) in compression, and A and B are

constants [14] (for a comprehensive review please see [15,16] ).

Studies of human cancellous bone [ 14,17

21 ] have reported the

exponent A to be as small as 1.2 [20] or as large as 3.0 [14] ,

suggesting that a bone specimen with 6% less bone mass due to

less bone volume is expected to be 7

–

17% less stiff. The

exponent B used to predict bone strength has been reported to be

as small as 1.48 [22] or as large as 2.47 [19] , suggesting that a

specimen with 6% less bone mass is expected to be 9

–

1 The term “ bone mass ” will be used here to describe the total mass of

mineralized tissue in a bone specimen (g) and should not be confused with

aBMD, a measure of bone density performed using dual-energy X-ray

absorptiometry that is expressed in the unit g/cm 2 .

–

14% less

strong ( Table 1 ).

Removal of trabeculae

–

where E is the stiffness of the specimen (elastic modulus),

1016

C.J. Hernandez / Bone 42 (2008) 1014 – 1020

Tissue degree of mineralization

applied this statistical approach to human bone and found the

following relationships [41] :

After a new volume of bone is formed, it begins to accu-

mulate mineral in a process that can continue for years after-

wards [23,24] . As a result, the degree of mineralization of older

bone tissue is greater than that of newly formed tissue, so that

the amount of bone turnover can influence the average tissue

degree of mineralization [25,26] .

Examination of mineralized tissues across a wide range of

species (including deer and whales) has associated increased

tissue degree of mineralization with increased bone stiffness

and strength and, in some cases, increased brittleness (the term

‘

ð Þ~

2 : 58 F 0 : 02

;

r Ult MPa

Þ~

2 : 79 F 0 : 09

;

is the degree

of mineralization (measured as ash mass / dry bone mass), and

the exponents are expressed as mean± standard error. This

analysis is the only study of cancellous bone to detect and

quantify the independent effects of bone volume and average

tissue degree of mineralization on bone biomechanics. With

regard to bone strength (Eq. (4)), the exponent applied to tis-

sue degree of mineralization (2.79) is much greater than the

exponent applied to bone volume fraction (1.92), suggesting

that cancellous bone strength is much more sensitive to

differences in tissue degree of mineralization. It is possible,

however, that clinical measures of bone mass might capture

some of this effect. Clinical measures of bone mass evaluate the

total amount of mineral present (both mineralized volume and

degree of mineralization). Assuming clinical evaluation of bone

mineral content (BMC) is directly related to the inorganic

content in bone (the ash content), BMC can be expressed as:

σ Ult and BV/TV are defined as above,

is used here in an engineering sense expressing a

material property) [27,28] . While these studies demonstrate

that tissue degree of mineralization can be biomechanically

important, most do not include analyses of bone porosity and

therefore cannot be used to examine the biomechanical effects

of tissue degree of mineralization independent of bone volume.

In addition, it is important to keep in mind that trends observed

among species do not necessarily apply within a species. For

example, consider the commonly held idea that

’

“

hyperminer-

bone tissue is more brittle, a concept frequently noted

when discussing possible adverse effects of long-term inhibi-

tion of bone turnover. Although comparisons among animals

suggest that more highly mineralized tissue is more brittle,

only two studies of human bone specimens have shown in-

creased tissue degree of mineralization to be associated with

increased brittleness (evaluated as impact energy in cortical

bone specimens) [29,30] . Indeed, other studies of human bone

specimens have found increased tissue degree of mineralization

to be associated with reduced brittleness (measured as com-

pressive toughness in cancellous bone [31] , or fracture

toughness evaluated in cortical bone [32] ). Additionally, a

number of studies of human bone specimens did not observe a

relationship between tissue degree of mineralization and

brittleness (measured as toughness or energy to failure in

cortical or cancellous bone [33

”

BMC

Þ q t a;

ρ t is the density of the mineralized bone tissue (in grams),

a parameter that is linearly related to tissue degree of mine-

ralization [13,41] . By combining Eqs. (3), (4) and (5) we can

estimate the differences in bone biomechanics associated with a

6% difference in BMC caused entirely by reductions in average

tissue degree of mineralization (this change in BMC corre-

sponds to a reduction in tissue degree of mineralization from

65% to 62% ash by weight). This difference in BMC is

associated with a difference in specimen stiffness of 11

–

13%

36] ). While comparisons among

species suggest that highly mineralized bone specimens are

more brittle, it is not yet clear that human bone can become

brittle through an increase in average tissue degree of mineral-

ization alone.

With regard to bone stiffness and strength, few studies have

been designed to separate the biomechanical effects of tissue

degree of mineralization from those of bone volume. Follet et al.

found that average tissue degree of mineralization (measured

through quantitative contact radiography) was positively cor-

related with cancellous bone stiffness, strength and brittleness

(brittleness evaluated as toughness) [31] , and that tissue degree

of mineralization had a biomechanical effect independent of

bone volume. Others have used power law models to predict

the biomechanical effects of tissue degree of mineralization in

cortical bone [18,37

–

13% (both of these

ranges are expressed using the 95% confidence interval of the

regression coefficients above).

Another way of comparing the biomechanical effects of bone

volume and tissue degree of mineralization is to examine the

relationship between bone mineral content and bone strength

[10] . Fig. 1 shows the percent change in bone strength expected

from a hypothetical reduction in bone mineral content under two

different conditions: 1) when changes in bone mineral content

are caused entirely by bone volume (lower region with dark blue

shading); and 2) when changes in bone mineral content are

caused entirely by tissue degree of mineralization (upper region

with lighter orange shading). That these two confidence intervals

overlap suggests that alterations in tissue degree of mineraliza-

tion may not modify the relationship between bone strength and

clinical measures of BMC. Hence, alterations in average tissue

degree of mineralization may have little effect on the ability of

BMC to predict bone strength and are therefore not expected to

be responsible for discrepancies between fracture incidence and

aBMD. Additional studies are needed to confirm this analysis

(Eqs. (3) and (4)) and to determine if the average degree of

–

40] , but only two of these studies

accounted for variation in bone volume fraction (both using

non-human tissue [38,40] ). Currey suggested that the separate

effects of bone volume and tissue degree of mineralization

could be expressed with a two-parameter power law model and

applied the approach to non-human tissue [38] . Hernandez et al.

–

E GPa

where E,

brittle

alized

where

and a difference in bone strength of 11

C.J. Hernandez / Bone 42 (2008) 1014 – 1020

1017

coefficient of variation of tissue stiffness from 20% to 50% is

expected to cause a 14% reduction in cancellous bone elastic

modulus.

Disconnection of trabeculae

9] . Quantifying the

effect of trabecular disconnection experimentally is challenging

because of technical difficulties in identifying and counting

individual trabeculae (only recently have techniques for directly

counting individual trabeculae in micro-CT images of cancel-

lous bone been presented [47,48] ). Existing biomechanical

analyses have therefore used cellular solid models with tra-

becular-like microarchitectures to mimic cancellous bone struc-

ture. Two- and three-dimensional cellular solid models indicate

that removal of individual trabeculae can result in reductions

in cancellous bone stiffness and strength that are greater than

would be expected from the associated change in bone volume

[49

–

Fig. 1. The predicted reduction in strength caused by a reduction in bone mineral

content or bone mineral density (as would be measured by dual-energy X-ray

absorptiometry) is shown. 1) a hypothetical case where changes in bone mineral

content are caused entirely by changes in volume fraction (dark blue region) and

2) a hypothetical case where changes in bone mineral content are caused entirely

by reductions in tissue degree of mineralization (light orange region). The

overlapping regions are based on the 95% confidence interval of the regression

model reported by Hernandez et al. [41] .

51] . Three-dimensional models suggest that a 6% differ-

ence in bone volume caused by removal of trabeculae can

reduce cancellous bone stiffness by 3% (only horizontal trabe-

culae removed) to 39% (only vertical/oblique trabeculae re-

moved) and compressive strength by 18% (only horizontal

trabeculae) to 35% (only vertical/oblique trabeculae) [51] .As

these simulations represent extreme cases where only horizontal

or only vertical/oblique trabeculae are removed, the actual

changes in bone biomechanics resulting from trabecular dis-

connection are expected to be somewhere in between these

values.

mineralization can have disproportionate effects on other me-

chanical properties (brittleness for example) or under different

loading conditions (impact, shear, etc.).

While most studies have concentrated on the biomechanical

effects of average tissue degree of mineralization, variation of

tissue degree of mineralization at the micro-scale has also been

implicated as a factor that can influence bone biomechanics.

Byaltering the number and/or size of new remodeling events,

bone turnover will not only modify the average tissue degree

of mineralization, but also the variability of tissue degree of

mineralization. Changes in variation of tissue degree of

mineralization have been associated with alterations in bone

turnover during bisphosphonate therapy and in metabolic bone

disease [42,43] . Variability of tissue degree of mineralization

can also differ among regions of the skeleton (iliac crest v.

calcaneous) [31] .

Micro-computed tomography based finite element models

have been useful for studying the biomechanical consequences

of variation in tissue degree of mineralization because they

make it possible to consider the biomechanical effects of tissue

degree of mineralization independent of bone volume or micro-

architecture. Finite element studies suggest that an increase

in intraspecimen variation in tissue degree of mineralization

(modeled as local variation in tissue stiffness) can cause a

reduction in cancellous bone stiffness, even when trabecular

microarchitecture and average tissue degree of mineralization

are maintained constant [44

Remodeling cavities

It has been proposed that cavities formed during bone remo-

deling (Howship's lacunae) can act as stress risers, causing

disproportionate reductions in the biomechanical performance

of cancellous bone. Experimental evaluation of the effects of

remodeling cavities on cancellous bone has been limited be-

cause a repeatable technique for making three-dimensional

measures of remodeling cavities in cancellous bone has not

yet been demonstrated. Existing data is therefore limited to

predictions made from finite element models. A number of

biomechanical analyses have illustrated how the presence of

a cavity on the surface of a single trabecula may increase the

stresses and strains in surrounding tissue [52

–

46] . Jaasma et al. determined that

an increase in the coefficient of variation (standard deviation/

mean) of the tissue stiffness from 20% to 50% would result in a

reduction in elastic modulus of cancellous bone by 19

5 mm in smallest

dimension [53,55] . A reduction in bone volume of 6% caused

by the addition of remodeling cavities was predicted to reduce

the elastic modulus by 12

–

24%,

even when average tissue degree of mineralization was main-

tained constant. More recently, Bourne and van der Meulen

measured variation in mineral content directly using calibrated

micro-computed tomography and found that an increase in the

–

47% and the compressive strength

61%. The ranges for these predictions are large because

the biomechanical effects of remodeling cavities can be

influenced by a number of factors including the initial bone

volume fraction (more porous bone can be more sensitive to

–

It is commonly stated that cavities formed during remodeling

can disconnect or fenestrate trabeculae, modifying trabecular

microarchitecture and potentially cause a disproportionate

change in cancellous bone strength [7

–

54] . Additionally,

two finite element analyses have suggested that remodeling

cavities can have a disproportionate effect on cancellous bone

stiffness and strength in specimens 3

–

by 13

1018

C.J. Hernandez / Bone 42 (2008) 1014 – 1020

remodeling cavities) and the placement of remodeling cavities

in the cancellous bone structure. When placed in regions of

high strain within the cancellous bone structure (where tissue

microdamage and mechanical stresses are expected to be

greatest) remodeling cavities can have a large, dispropor-

tionate effect on cancellous bone biomechanics [55] .Asa

result, the degree to which remodeling cavities are targeted to

tissue damage or tissue strain (two factors believed to stimulate

bone remodeling) will modulate the effect of bone remodeling

on bone biomechanics.

Additionally, the number and size (length, width, depth) of

remodeling cavities may have biomechanical significance.

Although an increase in bone turnover is commonly interpreted

as an increase in the number of remodeling events, two-dimen-

sional histomorphometry measurements cannot differentiate

an increase in the number of remodeling events from an increase

in the size of each individual event (width, length and depth)

[56] , a distinction that can result in very different stress dis-

tributions within cancellous bone. Simple mechanical analyses

suggest that the number and size of remodeling cavities may

influence the mechanical performance of a trabecula indepen-

dent of bone volume or total amount of bone turnover ( Fig. 2 ).

The number and size of remodeling cavities may also influence

intraspecimen variation in tissue degree of mineralization by

determining the size of each osteon or hemi-osteon and may

also influence the rate at which trabeculae are disconnected by

remodeling events (deeper cavities are more likely to disconnect

trabeculae [57] ). Unfortunately, little is known about the com-

plete size and shape (length, width and depth) of remodeling

cavities in human cancellous bone because two-dimensional

techniques cannot obtain measures of all three of these size

dimensions at once [58,59] . Micro-computed tomography is not

as helpful as one would expect because few imaging systems

can obtain the resolution needed to detect the scalloped surface

of a remodeling cavity. Those imaging systems with such high

resolution have a very limited field of view such that only one or

two cavities can be viewed per specimen, far too few to

characterize the population of remodeling events in cancellous

bone. Recently, serial block-face imaging using an automated

microtome or milling machine has been used to image

remodeling cavities in three dimensions, and may prove useful

in determining the placement and size of remodeling cavities in

human bone biopsies or cadaver tissue [54,60] .

Conclusions

Table 1 provides a unique way of comparing the biomecha-

nical effects of bone remodeling for a given difference in bone

mass (in this case a 6% difference in bone mass). An aspect of

bone that has the potential to explain discrepancies between

aBMD and fracture incidence will have a biomechanical effect

that is greater than would be expected from that caused by bone

volume alone. For example, a 6% difference in bone mass

caused by the average tissue degree of mineralization (second

row in Table 1 ) is associated with an 11

–

14%,

first row of Table 1 ). As a result, this analysis suggests that it is

unlikely that differences in average tissue degree of mineraliza-

tion can have a disproportionate effect on cancellous bone

compressive strength, and that average tissue degree of

mineralization would be unlikely to contribute to a biomecha-

nical explanation for discrepancies between aBMD and fracture

incidence.

Two conclusions can be made from comparing the remaining

factors in Table 1 to the effect of bone volume. First, existing

experimental and computational data suggest that, while aver-

age tissue degree of mineralization can influence bone strength,

it may not be able to explain discrepancies between bone

biomechanics and clinical measures of bone mass. Local vari-

ability of tissue degree of mineralization is a more likely

explanation. Secondly, while existing biomechanical analyses

support the idea that trabecular disconnection and remodeling

cavities may have a disproportionate effect on bone biomecha-

nics (i.e. the biomechanical effects can exceed those expected

from difference in bone volume), there is a considerable overlap

between the biomechanical effects of these factors and the

biomechanical effects of differences in bone volume alone. As a

result, further work is needed to determine if bone remodeling

may have a disproportionate effect on bone biomechanics.

Whether or not these aspects of bone remodeling have a

biomechanically relevant effect independent of bone mass will

depend on characteristics that we currently know little about,

such as the number and size of remodeling events and how well

remodeling cavities are targeted to mechanical stress/strain and

microscopic tissue damage. Additionally there is growing evi-

dence that the concentration of naturally occurring non-

–

Fig. 2. Three cylindrical trabeculae (150 μ m in diameter) are shown with remodeling cavities wrapped around them circumferentially. The remodeling cavities in each

image occupy the same volume (i.e. the same amount of bone turnover is depicted) but the number, surface size and depth of the cavities differs within the range

observed histologically. The bending moment and critical load for Euler buckling (calculated within the tapered region only) is presented relative to that in the first

image and is shown to vary by as much as an order of magnitude among the three possibilities.

13% difference in bone

strength, a range that is well within what is expected for the

same reduction in bone mass caused by bone volume (9

Plik z chomika:

Inne pliki z tego folderu:

Inne foldery tego chomika: