1

Pathogenesis of Fragility Fractures:A Biomechanical View

Mary L. BouxseinOrthopaedic Biomechanics LaboratoryBeth Israel Deaconess Medical CenterBoston, MAUSA

2

Outline

• Pathogenesis of fractures and determinants of bone strength

• Age-related changes that contribute to skeletal fragility

• Interaction between skeletal loading and bone strength

• Non-invasive assessment of bone strength

3

Design of a structure

• Consider what loads it must sustain

• Design options to achieve desired function– Overall geometry

– Building materials

– Architectural details

4

Determinants of whole bone strength

• Geometry– Gross morphology (size & shape)

– Microarchitecture

• Properties of Bone Material / Bone Matrix– Mineralization

– Collagen characteristics

– Microdamage

5

Hierarchical nature of bone structure

Seeman & DelmasN Engl J Med 2006; 354:2250-61

Macrostructure

Microstructure

Matrix Properties

Cellular Composition and Activity

6

FRACTURE?

Loads applied to the bone

Bone strength

Hayes et al. Radiol Clin N Amer. 1991; 29:85-96Bouxsein et al. J Bone Miner Res. 2006; 21:1475-82

Applied loadBone strength

> 1 fractureFactor ofrisk

Bone MassGeometry

Material properties

Fall traitsProtective responses

Bending, lifting

Biomechanical approach to fractures

7

Pathogenesis of fragility fractures

Age-related changes that contribute to fragility fractures:

1) Decreased bone strength

2) Increased propensity to fall

8

Assessing bone biomechanical properties

Deformation

Force

StructuralProperties

MaterialProperties

9

Biomechanical testingKey properties

Displacement

Force

Failure load

Energy absorbed

Stiffness

10Review of the relevant material & geometric properties vs loading mode

compression

bending

Cross-sect area, A ∝ r2

Axial rigidity, E•A

Moment of inertia, I ∝ r4

Section modulus, Z ∝ r3

Bending rigidity, E•I

Material: Elastic modulus E ∝ (density)a

Ultimate strength S ∝ (density)b

11

Effect of cross-sectional geometry on long bone strength

aBMD (by DXA) = = ↓

Compressive strength

Bending strength

=

=

↑

↑↑

↑

↑↑↑

Bouxsein, Osteoporos Int, 2001

12

Bone strength

SIZE & SHAPE macroarchitecturemicroarchitecture

MATERIAL tissue compositionmatrix properties

BONE REMODELINGformation / resorption

OSTEOPOROSIS DRUGS

Bouxsein. Best Practice Res Clin Rheumatol, 2005; 19:897-911

13

Outline

• Determinants of bone strength

• Age-related changes that contribute to skeletal fragility

14

Age-related changes in mechanical properties of bone tissueCortical bone

% loss 30-80 yrs

-8%

-11%

-34%

Elastic modulus, E

Ultimate strength, S

Toughness

Cancellous bone% loss 30-80 yrs

-64%

-68%

-70%

Bouxsein & Jepsen, Atlas of Osteoporosis, 2003

15Whole bone strength declines

dramatically with age

0

2000

4000

6000

8000

10000

young

old

Femoral neck(sideways fall)

Courtney et al. J Bone Joint Surg Am. 1995; 77:387-95 Mosekilde. Technology and Health Care 1998; 6:287-97

Lumbar vertebrae(compression)

Who

le b

one

stre

ngth

(New

tons

)

0

2000

4000

6000

8000

10000

young old

16Age-related changes in geometry

Adaptation to maintain whole bone strength

Seeman. Lancet 2002; 359: 1841-50 Seeman. N Engl J Med. 2003; 349:320-3

17

Age-related changes in femoral neck vBMD, geometry, and strength indices

Trabecular vBMDCortical vBMD

Total areaMOIapAxial rigidityBending rigidity

WomenF vs M

(P-value)Men% Change, ages 20-90 yrs

- 56**- 24**

13**- 1- 39**- 25**

- 45**- 13**

7*- 21**- 31**- 24**

<0.001<0.001

nsnsnsns

For age regressions: *P<0.05, **P<0.005

Riggs et al. J Bone Miner Res. 2004; 19:1945-54Riggs et al. J Bone Miner Res. 2006; 21(2):315-23

368 women, 320 men, aged 20-97

18

Age-related changes in trabecular microarchitecture

• Decreased bone volume, trabecular thickness and number

• Decreased connectivity

• Decreased mechanical strength

Image courtesy of David Dempster

19

Microarchitectural changes that influence bone strength

Force required to cause a slender column to buckle:

• Directly proportional to– Column material

– Cross-sectional geometry

• Inversely proportional to– (Length of column)2 www.du.edu/~jcalvert/tech/machines/buckling.htm

20

Theoretical effect of cross-struts on buckling strength

# Horizontal Effective Buckling Trabeculae Length Strength

0 L S

1 1/2 L 4 x S

Bouxsein. Best Practice Res Clin Rheumatol, 2005; 19:897-911

21

Cortical porosity increases with age(41 iliac biopsies, age 19-90)

Age (years)

0

3

6

9

12

15

0 20 40 60 80

r = 0.78 P < 0.001

(%)

4-fold increase incortical porosity from

age 20 to 80

Increased heterogeneity with age

Brockstedt et al. Bone 1993; 14:681-91

22

20-year-old 80-year-oldMayhew et al, Lancet 2005

Age-related changes in femoral neck cortex and association with hip fracture

Those with hip fractures have:

• Preferential thinning of the inferior anterior cortex• Increased cortical porosity

Bell et al. Osteoporos Int 1999; 10:248-57Jordan et al. Bone, 2000; 6:305-13

23

Age-related changes in bone properties associated with fracture risk

• Decreased bone mass and BMD

• Altered geometry

• Altered architecture– Cortical thinning

– Cortical porosity

– Trabecular deterioration

Images from L. Mosekilde, Technology and Health Care. 1998

young

elderly

24

Outline

• Determinants of bone strength

• Age-related changes that contribute to femoral fragility

• Interaction between skeletal loading and bone strength

25

Etiology of age-related fractures

FRACTURE?

Loads applied to the bone

Bone strength

26

Φ > 1, FrxΦ < 1, no Frx

Applied loadBone strength

The factor of risk concept

• Identify activities associated with fracture

• Use biomechanical models to determine loads applied to bone for those activities

• Estimate bone failure load for those activities

Φ =

Hayes et al. Radiol Clin N Amer. 1991; 29:85-96 Bouxsein et al. J Bone Miner Res. 2006; 21:1475-82

27

Falls and hip fracture

• Over 90% of hip fx’s associated with a fall

• Less than 2% of falls result in hip fracture

• Fall is necessary but not sufficient condition

• Sideways falls are most dangerous

• What factors dominate fracture risk?

28

Factor Adjusted Odds Ratio

Fall to side 5.7 (2.3 - 14)Femoral BMD 2.7 (1.6 - 4.6)*Fall energy 2.8 (1.5 - 5.2)**Body mass index 2.2 (1.2 - 3.8)*

* calculated for a decrease of 1 SD** calculated for an increase of 1 SD

Greenspan et al, JAMA, 1994; 271(2):128-33

Independent risk factors for hip fracture

29Estimating loads applied to the hip during a

sideways fall

Human cadavers

Human volunteers

Crash dummy

Mathematical models and simulations

Peak impact forces applied to greater trochanter: 270 - 730 kg (2400 - 6400 N)

(for 5th to 95th percentile woman)

Robinovitch et al. 1991; Biomech Eng. 1991; 113:366-74Robinovitch et al.1997; Ann Biomed Eng. 1997; 25:499-508

van den Kroonenberg et al J BiomechanicEngl.1995; 117:309-18 van den Kroonenberg et al J Biomech.

1996; 29:807-11

30

Femur is strongest in habitual loading conditions

Keyak et al. J Biomech. 1998; 31:125-33

0123456789

Stance Sideways fall

Failu

re lo

ad (k

N)

P < 0.0012318 ± 300 N7978 ± 700 N

31

Femur is strongest in habitual loading conditions

Keyak et al. J Biomech. 1998; 31:125-33 van den Kroonenberg et al. J Biomech. 1996; 29:807-11

0123456789

Stance Sideways fall

Failu

re lo

ad (k

N)

P < 0.0012318 ± 300 N7978 ± 700 N

Φ =Applied loadFailure load

Thus, Φ > 1 for sideways fall in elderly persons

Fall load

32

Vertebral fractures

• Difficult to study– Definition is controversial– Many do not come to clinical attention– Slow vs. acute onset– The event that causes the fracture is often

unknown

• Poor understanding of the relationship between spinal loading and vertebral fragility

33

Estimating loads on the lumbar spine

Schultz et al.1991; Spine. 1991; 16:1211-6 Wilson et al.1994; Radiology. 1994 ; 193:419-22 Bouxsein et al. J Bone Miner Res. 2006; 21:1475-82

34

Factor of risk for vertebral fracture (L2)Bending forwards 90o with 10 kg weight in hands

Age

0.00.20.40.60.81.01.21.41.61.8

10 20 30 40 50 60 70 80 90 100

Men

0.00.20.40.60.81.01.21.41.61.8

10 20 30 40 50 60 70 80 90 100Age

Women

11.9% 30.1%

+92% over life** †+28% over life**

** P<0.005 for age-regressions† p<0.01 for comparison of age-related change in M and W

Bouxsein et al. J Bone Miner Res. 2006; 21:1475-82

35

Proportion of individuals with

Φ > 1, per decade

0102030405060

20-29 30-39 40-49 50-59 60-69 70-79 80+

%

4%4% 11% 12%

21%11%

27%

42%

53%

Age

(Bending forwards 90o lifting 10 kg weight)

0200400600800

100012001400

40 50 60 70 8030 90

Incidence / 100,000 person-years

Bouxsein et al. J Bone Miner Res. 2006; 21:1475-82

36

Outline

• Determinants of bone strength

• Age-related changes that contribute to femoral fragility

• Interaction between skeletal loading and bone strength

• Non-invasive assessment of bone strength

37

Bouxsein et al, 1999

• Does not distinguish– Specific attributes of 3D geometry– Cortical vs cancellous density– Trabecular architecture– Intrinsic properties of bone matrix

Clinical assessment of bone strength by DXA• Areal BMD by DXA

– Bone mineral / projected area (g/cm2)

• Reflects (indirectly)– Geometry / Mass / Size– Mineralization

• Moderate to strong correlation with whole bone strength at spine, radius & femur (r2 = 50 - 90%)

• Strong predictor of fracture risk

38

Estimating hip geometry from 2D DXA“Hip Strength Analysis”

Estimate femoral geometry and strength indices• Use 2D image data to derive 3D geometry• Requires assumptions that have not been tested for all populations and

treatmentsBeck et al. 1999, 2001

39

QCT assessment of bone density and geometry

Images courtesy of Dr. Thomas Lang, UCSF

40

Non-Fx

62 yr

Fx

62 yr

Ito et al. J Bone Miner Res. 2005; 20:1828-36

Multi-detector computed tomography

Potential to show trabecular architecture…Higher resolution, but higher radiation dose

41

Trabecular architecture in vivowith high resolution pQCT

~ 80 µm3 voxel size

~ 3 min scan time, < 4 µSv

Distal radius and tibia only

Reproducibility: density: 0.7 - 1.5% *µ-architecture: 1.5 - 4.4% *

Xtreme CT, Scanco

* Boutroy et al. J Clin Endocrinol Metab. 2005; 90:6508-15

42

Premenopausal

PostmenopausalOsteopaenia

Postmenopausal Osteoporosis

PostmenopausalSevere Osteoporosis

Boutroy et al. J Clin Endocrinol Metab. 2005; 90:6508-15

Tibia Radius

43

Discrimination of osteopaenic women with and without history of fracture by HR-pQCT

(age = 69 yrs, n=35 with prev frx, n=78 without fracture)

* p < 0.05 vs fracture free controls

Boutroy et al. J Clin Endocrinol Metab. 2005; 90:6508-15

0.4%

-12%*-9%*

13%*

26%*

-0.3%

-10%

0%

10%

20%

30%

SpineBMD

Fem NeckBMD BV/TV TbN

TbSp TbSpSD

% D

iffer

ence

TbTh

-5%

44

MRI for architecture assessment in vivo

• Features of MRI– Non-invasive– No x-rays– True 3D – Oblique scanning plane– Clinical scanners– Image time: 12 - 15 minutes~ 150 x 150 x 300 µm,

60 x 60 x 100 µm

Image courtesy of Dr. Felix Wehrli, UPenn

Image courtesy of Dr. Sharmila Majumdar, UCSF

45

Images from two women with similar BMD

www.micromri.com

MRI assessment of trabecular structure

46

Crawford et al, Bone 2003; 33: 744-750

QCT-based finite element analysis

• FEA is a well-established engineering method for analysis of complex structures

• Integrates geometry and density information from QCT scan to provide measures of bone strength

• In some cases, more strongly associated with whole bone strength in cadavers than DXA

• Further clinical validation needed

Image courtesy of T. Keaveny

Faulkner et al, Radiology 1991; Keyak et al, J Biomechanics 1998; Pistoia, Bone 2002; van Rietbergen JBMR 2003; Crawford et al, Bone 2003

47

Conclusions

• Whole bone strength is determined by bone mass, geometry, microarchitecture and characteristics of bone material

• Hip fractures result from an increase in traumatic loading —in particular, falls to the side — coupled with a deterioration in bone strength with increased age

• Biomechanically based assessment of fracture risk may improve diagnosis and understanding of treatment effects

• New tools are being developed for non-invasive assessment of bone strength and fracture risk