Biomechanical characteristics of regenerated cortical bone in the canine mandible.

Abstract

To test the mechanical properties of regenerate cortical bone created using mandibular bone transport (MBT) distraction, five adult male American foxhound dogs underwent unilateral distraction of the mandible with a novel MBT device placed to linearly repair a 30-35 mm bone defect. The animals were sacrificed 12 weeks after the beginning of the consolidation period. Fourteen cylindrical specimens were taken from the inner (lingual) and outer (buccal) plates of the reconstructed mandible and 21 control specimens were removed from the contralateral aspect of the mandible. The mechanical properties of the 35 cylindrical cortical bone specimens were assessed by using a non-destructive pulse ultrasound technique. Results showed that all of the cortical mechanical properties exhibit higher numerical values on the control side than the MBT regenerate side. In addition, both densities and the elastic moduli in the direction of maximum stiffness of the regenerate cortical bone specimens are higher on the lingual side than the buccal side. Interestingly, there is no statistical difference between elastic modulus (E(1) and E(2) ) in orthogonal directions throughout the 35 cortical specimens. The data suggest that not only is the regenerate canine cortical bone heterogeneous, but the elastic mechanical properties tend to approximate transverse isotropy at a tissue level, as opposed to control cortical bone, which is orthotropic. In addition, the elastic mechanical properties are higher not only on the control side but also in the lingual anatomical position, suggesting a stress shielding effect from the presence of the reconstruction plate.

Full text

Page 1

JOURNAL OF TISSUE ENGINEERING AND REGENERATIVE MEDICINE
J Tissue Eng Regen Med (2011).
Published online in Wiley Online Library (wileyonlinelibrary.com) DOI: 10.1002/term.347
RESEARCH ARTICLE
Biomechanical characteristics of regenerated
cortical bone in the canine mandible
Uriel Zapata1,2,3, Lynne A. Opperman1, Elias Kontogiorgos1, Mohammed E. Elsalanty4
and Paul C. Dechow1*
1Texas A&M Health Science Center, Baylor College of Dentistry, Dallas, TX, USA
2Eafit University, Mechanical Engineering Department, Medell´ ın, Colombia
3Division of Basic Medical Sciences, Mercer University School of Medicine, Macon, GA, USA
4Medical College of Georgia, Augusta, GA, USA
Abstract
To test the mechanical properties of regenerate cortical bone created using mandibular bone
transport (MBT) distraction, five adult male American foxhound dogs underwent unilateral
distraction of the mandible with a novel MBT device placed to linearly repair a 30–35 mm
bone defect. The animals were sacrificed 12 weeks after the beginning of the consolidation period.
Fourteen cylindrical specimens were taken from the inner (lingual) and outer (buccal) plates
of the reconstructed mandible and 21 control specimens were removed from the contralateral
aspect of the mandible. The mechanical properties of the 35 cylindrical cortical bone specimens
were assessed by using a non-destructive pulse ultrasound technique. Results showed that all
of the cortical mechanical properties exhibit higher numerical values on the control side than
the MBT regenerate side. In addition, both densities and the elastic moduli in the direction of
maximum stiffness of the regenerate cortical bone specimens are higher on the lingual side than
the buccal side. Interestingly, there is no statistical difference between elastic modulus (E1 and
E2) in orthogonal directions throughout the 35 cortical specimens. The data suggest that not only
is the regenerate canine cortical bone heterogeneous, but the elastic mechanical properties tend
to approximate transverse isotropy at a tissue level, as opposed to control cortical bone, which is
orthotropic. In addition, the elastic mechanical properties are higher not only on the control side
but also in the lingual anatomical position, suggesting a stress shielding effect from the presence of
the reconstruction plate. Copyright  2011 John Wiley & Sons, Ltd.
Received 2 March 2010; Accepted 8 July 2010
Keywords
healing; jaw mechanics
ultrasound technique; animal model; bony tissue; bone transport; distraction device; bone
1. Introduction
Bone transport distraction osteogenesis is a surgical
technique in which a bony disc is moved within
a segmental defect to fill the gap with new bony
tissue (Sacco and Chepeha, 2007). Over the past
decade, mandibular bone transport (MBT) has been an
alternative procedure to repair mandibular segmental
defects produced by trauma and oncology resections
*Correspondence to: Paul C. Dechow, 3302 Gaston Avenue,
Department of Biomedical Sciences, Dallas, TX 75246, USA.
E-mail: PDechow@bcd.tamhsc.edu
(Kuriakose et al., 2003). MBT is able to achieve
mandibular reconstruction in a predictable way, with few
complications (Imola et al., 2002), less complexity than
other procedures (Rubio-Bueno et al., 2000) and minimal
tissue morbidity (Karp et al., 1990). However, assessing
the biomechanical quality of the newly formed cortical
bone after MBT is important to evaluate the success of
the distraction process (Li et al., 2002), to define the best
time for the endpoint of the consolidation period (Claes
and Cunningham, 2009), to determine the optimal time
for the removal of the MBT device (Orbay et al., 1992)
and also to better understand the relationship between
mechanical properties and microstructure (Robertson and
Copyright  2011 John Wiley & Sons, Ltd.

Page 2

U. Zapata et al.
Smith, 1978; Lipson and Katz, 1984). This knowledge
will help to advance understanding the biomechanical
behaviour of the mandible after MBT by supporting the
definition of accurate numerical models gained from the
finite element method (FEM) (Casta˜ no et al., 2002; de
Zee et al., 2007).
Although, to our knowledge, no previous studies have
reported the mechanical properties of the new bony tissue
obtained after MBT, several attempts have been made
to determine the differences in mechanical properties
between newly formed bone after monofocal distraction
osteogenesis and mature bone. The first report of canine
mechanical properties after distraction osteogenesis was
presented by Orbay et al. (1992); they examined the
torsional properties of canine tibial bones at 4, 6, 8 and
10 weeks after the completion of monofocal distraction
and found great variability in mechanical properties
over time. The same year, Schickendantz et al. (1992)
tested canine tibial bones after monofocal distraction
by using torsional load tests and found that several
regenerate bone samples were as strong as the control
tibial bones. The first distracted mandibular bone strength
test was performed in a unilateral bifocal distraction
experiment by Costantino et al. (1993); they performed a
three-point bending test of the mandibular bone from
three mongrel dogs at four different sites: a control
reference from the contralateral untreated hemimandible,
the docking site, the transport disc and the newly grown
bone, and found the average ultimate strength of the
regenerate specimens was 77% of the strength of the
control specimens after 56 weeks of consolidation and
mineralization. However, due to the small sample size
and high variability, the results were not significant.
Walsh et al. (1994) suggested that mechanical properties
of the regenerate bone increased with time until reaching
50% of the tensile properties of the original cortical
bone in the canine tibia after a 12 weeks consolidation
period. Fitch et al. (1996) evaluated the biomechanical
properties of regenerate canine tibial cortical bone using
a cantilever-bending test; they found that the bending
stiffness of the new bone by day 79 (13 weeks) was
37–46%, which was significantly less than the control
bone. Kunz et al. (2006) evaluated mechanical properties
of the regenerated bone by performing a compression
test of mandibular cortical bone specimens obtained
after a bilateral distraction process in a canine model;
They found that the regenerate bone was 62.5% as
stiff as the control bone after a 13 week consolidation
period.
Although all the previous reported attempts to search
for differences between regenerate cortical bone and
control cortical bone after mandibular distraction osteo-
genesis have assumed the cortical bone to be an isotropic
material (mechanical properties present no variability by
direction) (Ashman et al., 1985), we have assumed the
anisotropic possibility for regenerate and control cor-
tical bone by using a ultrasonic wave technique that
enables determination of elastic coefficients in three
dimensions (3D) by measuring the multiple velocities of
ultrasonic waves through cylindrical cortical specimens in
controlled directions to characterize their elastic proper-
ties (Ashman et al., 1984; Schwartz-Dabney and Dechow,
2003).
The purpose of this study was to test the hypotheses
that the elastic properties in canine mandibular regen-
erate and control cortical bone at a tissue level are
(1) heterogeneous relative to each other, but (2) can both
be approximated as similarly orthotropic.
2. Materials and methods
Five adult male American foxhound dogs, weighing
60–94 lb, approximately 2 years old, underwent MBT
by creating a 30–35 mm unilateral mandibular defect.
The jaw was then stabilized with a novel reconstruction
plate to which an internal transport distraction unit was
attached (Figure 1). After 7 days of latency, the device
was activated at a rate of 1 mm/day for 4–5 weeks. Half-
way through the distraction period, molars were extracted
on the contralateral control side of the mandible, creating
an approximately 30 mm long edentulous control ridge.
Once distraction was accomplished on the experimental
side, the regenerate bone was allowed to consolidate for
6 weeks before placement of dental implants on both the
experimental and control sides. After a healing period of
6 weeks, the animals were sacrificed and their mandibles
were dissected. The mandibles were stored at −20◦C
prior to removal of bone specimens to ensure that the
bone’s elastic properties were preserved (Zioupus and An,
2000).
Thirty-five cylindrical cortical bone specimens (mean
diameter = 5.93 mm, SD = 0.06 mm; cortical thickness =
2.72 mm, SD = 0.24 mm) were removed from the lingual
and buccal cortices of the regenerate and control sides
of the mandible through the use of a low-speed dental
drill continuously cooled with a water drip. Twenty-one
cylindrical cortical specimens were taken from the control
sides, and 14 from the regenerate sides (Table 1). Prior to
Figure 1. Radiographic image from a reconstructed dog’s
mandible. The MBT device is positioned on the buccal side,
with the reconstruction plate (arrowheads) spanning the new
bone gap (asterisks) present after transport, and the transport
unit (arrow) attached to the bone disk at the anterior border of
the gap
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term

Page 3

Mechanical properties of canine cortical bone
Table 1. Regenerate and control cylindrical cortical bone speci-
mens: distribution among subjects, relative positions (alveolar,
medial and basal) and anatomical positions (buccal and lingual)
Regenerate boneControl bone
SubjectPosition BuccalLingual Buccal Lingual
Dog 1Alveolar
Medial
Basal
Alveolar
Medial
Basal
Alveolar
Medial
Basal
Alveolar
Medial
Basal
Alveolar
Medial
Basal
−
−
−
×
−
×
−
−
−
×
−
×
×
−
−
5
×
−
×
×
−
×
−
×
−
×
−
×
×
−
×
9
×
×
×
×
−
×
×
−
×
×
−
×
×
−
×
11
×
×
×
×
−
×
−
−
×
×
−
×
×
−
×
10
Dog 2
Dog 3
Dog 4
Dog 5
Subtotal
removal, each bone specimen was marked with a graphite
line parallel to the lower border of the mandibular corpus,
indicating the anatomical orientation. After extraction, all
the cortical samples were stored in a 50:50 solution of
95% ethanol and isotonic saline at room temperature
(19◦C) to maintain the elastic properties of cortical bone
with minimal changes over time (Ashman et al., 1984).
Cortical bone apparent density (mg/cm3) was calculated
from sample weights taken with an analytical balance
(PM460, Mettler-Toledo International INC, Columbus,
OH, USA) according to Archimedes’ principle of buoyancy
(Ashman et al., 1984).
Three orthogonal principal axes of stiffness were
defined within each cortical specimen by placing the
cylindrical cortical specimens between two ultrasonic
longitudinal transducers (Olympus V312, immersion,
0.25 diameter; Waltham, MA, USA) that sent 10 MHz
ultrasonic pulses through the specimens (Schwartz-
Dabney and Dechow, 2003). Ultrasonic velocities were
calculated by dividing either the specimen thickness or
specimen diameter by the time delay. The time delay is
the difference between the start of the transmitted wave
applied to one side of the specimen and the start of the
received ultrasonic signal atthe other side of the specimen
(Ashman et al., 1984). The specimens were tested in nine
consecutive directions around the circumference at 22.5◦
intervals, starting at the graphite reference line. Then
three orthogonal principal directions were defined for
each cylindrical specimen with the following rule: 3 is
the direction of maximum stiffness in the plane of the
cortical plate, 2 is the direction of minimum stiffness in
the same plane, and 1 is the direction perpendicular to
the cylindrical plane of the specimen (Schwartz-Dabney
and Dechow, 2003).
Elastic mechanical properties of the cortical bone were
calculated from measured longitudinal and transverse
(Olympus V156, 5 MHz, contact, 0.25 diameter) ultra-
sonic wavevelocities usingthe relationship betweenstress
and strain tensors as defined in Hooke’s law (Ashman
et al., 1984; Rho, 1996). Notation for how the elastic
mechanicalpropertieswere relatedtotheprincipalaxesof
stiffness is as follows: the elastic modulus E is followed by
a subscript indicating the corresponding principal direc-
tion (E1, E2, E3); the shear modulus G is followed by two
subscripts indicating the principal plane of shear (G13,
G12, G23); and the Poisson’s ratios υ are denoted by a
double subscript indicating the principal direction of load
(first subscript) followed by the orthogonal direction for
Poisson’s strain (second subscript)(υ12, υ21, υ23, υ32, υ13,
υ31) (Marinescu et al., 2005). The method of calculation
assumed that the cortical bone specimens are basically
orthotropic in structure, a conjecture supported by the
symmetry of the curves generated when plotting variation
in ultrasonic velocities by orientation within each speci-
men. A consequence of orthotropy is that υijEi= υji Ej
and thus only three Poisson ratios need to be reported:
υ12, υ13, and υ23. In addition, anisotropy was quantified
in the cortical plane by using ratios of principal elastic
moduli (E2/E3).
Data were analysed using SPSS statistical software
(SPSS Inc., Chicago, IL, USA) to test for differences in
mechanical properties between regenerate and control
mandibular cortical bone. ANOVAs were used to test for
differences among the five dogs. Because cortical bone
specimens were removed from consistent locations on
both sides of the mandible, paired t-tests were used to test
for differences between the buccal and lingual sides of the
mandibles. Five pairs of specimens from the regenerate
side of the mandible, representing three animals, and 10
pairs from the control side, representing five animals,
were compared (Table 1). Paired t-tests were also used
to test for a difference between superior and inferior
relative positions in six paired regenerate specimens from
four animals, and nine paired control specimens from five
animals (Table 1).
The orientation of maximum stiffness was assessed
by using the circular statistical software Oriana (Kovach
Computing Services. Anglesey, Wales). Tests included
Rayleigh’s uniformity test to determine the presence
of a common orientation among specimens, as well as
Watson–Williams tests to determine differences between
groups.
3. Results
All the five dogs survived the entire MBT process without
major surgical complications. The healing process was
normal among the dog subjects without affecting feeding
or physical activity. However, the third dog experienced
detachment of the new tissue, which was immediately
reattached; and the fourth dog showed necrosis of the
transport disc, although new bony tissue was regenerated.
MBT disc advance was performed without complications
or apparent animal discomfort. All the subjects reached
MBT reconstruction with adequate bony volume and
external shape.
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term

Page 4

U. Zapata et al.
3.1. Density
The mean density of all the 35 cylindrical cortical
bone specimens was 1883.5 ± 173.7 mg/cm3(Table 2).
No statistical differences were found among animals in
density (F = 0.353, p = 0.84). Likewise, no significant
differences in density were found in the control cortical
bone between buccal and lingual sides. However, a
significant difference in density was found in the
regenerate bone (p = 0.009) between buccal and lingual
sides (Figure 2A). In comparisons between mandibular
locations, no significant differences in the regenerate
cortical bone were found between the alveolar and
basal positions, whereas, in the control cortical bone
the difference between the alveolar and basal positions
(Figure 2B) was significant (p = 0.045). Overall, density
was significantly less (F = 24.55; p < 0.001) in the
regenerate cortical bone (1746.6 ± 183.2 mg/cm3) than
in the control cortical bone (1974.8 ± 87.2 mg/cm3;
Figure 2C).
3.2. Maximum stiffness orientation
No significant differences in maximum stiffness orienta-
tion were found between regenerate and control cortical
bone, alveolar and basal positions, or lingual and buccal
sides (Figure 3A). The principal averaged circular mean
direction for the 35 specimens was 12.4◦with a circular
standard deviation (SD) of 12.0◦, and length of mean
vector (r) of 0.978 (Figure 3B). Since r is a measurement
of circular dispersion, a value close to 1 represents a large
concentration of the data around the mean.
3.3. Elastic moduli
The elastic modulus E2 expressed statistical differences
between dogs 1 and 5 (p = 0.031); complementarily,
E1 denoted statistical differences between dogs 1 and
3 (p = 0.02) as well as dogs 1 and 5 (p = 0.015).
Although significant differences were found in elastic
moduli E1(p = 0.03), E2(p = 0.015) and E3(p = 0.017)
between buccal and lingual anatomical positions on
the regenerate cortical bone specimens (Figure 4A), no
significant differences were found in elastic moduli in
the control cortical bone specimens (Table 2). Testing
for differences between alveolar and basal positions, a
statistical difference (p = 0.01) was found for E3between
the alveolar (29.2 ± 2.5 GPa) and basal (33.2 ± 2.6 GPa)
relative positions (Table 2).
Significant differences in elastic modulus E1 (F =
11.12, p = 0.002), E2 (F = 15.04, p = 0.001) and E3
(F = 23.65, p = 0.001) were found between regenerate
and control bone specimens (Figure 4B). Furthermore,
the statistical evaluation of all 35 specimens showed
significant differences in principal elastic moduli between
directions 1 and 3 (p = 001) and 2 and 3 (p = 0.001). In
contrast, no significant differences were found between
directions 1 and 2 (p = 0.051) (Figure 4C).
Table 2. Elastic mechanical properties of the canine cortical bone determined by ultrasound wave technique for regenerate and
control cortical bone for two anatomical positions (buccal and lingual) and two relative position on the mandible (alveolar and basal)
Cortical
bone
Specimen
position
Density
(δ)
(mg/cm3)
E1
(GPa)
E2
(GPa)
E3
(GPa)
G12
(GPa)
G31
(GPa)
G23
(GPa)
υ12
υ13
υ23
RegenerateBuccal
Lingual
Buccal
Lingual
Alveolar
Basal
Alveolar
Basal
1557.7 ± 56.5
1861.1 ± 142.0 10.7 ± 1.9 13.1 ± 2.1 27.5 ± 6.5 4.0 ± 0.7 6.2 ± 1.4 7.9 ± 1.5 0.45 ± 0.06 0.18 ± 0.06 0.08 ± 0.03
1986.2 ± 42.7
1971.6 ± 119.0 12.7 ± 2.5 15.7 ± 2.7 31.3 ± 3.7 4.8 ± 1.0 7.4 ± 1.2 9.2 ± 1.4 0.42 ± 0.06 0.15 ± 0.02 0.12 ± 0.03
1736.8 ± 189.5
1819.8 ± 168.5 11.2 ± 2.6 12.7 ± 3.5 26.5 ± 7.9 4.0 ± 1.1 6.3 ± 1.5 7.6 ± 1.9 0.47 ± 0.05 0.19 ± 0.07 0.09 ± 0.07
1921.9 ± 104.9 11.6 ± 1.0 14.4 ± 0.7 29.2 ± 2.5 4.4 ± 0.3 6.6 ± 0.7 8.2 ± 0.7 0.41 ± 0.05 0.15 ± 0.02 0.13 ± 0.03
2023.7 ± 41.9
7.8 ± 0.9
12.6 ± 2.0 15.7 ± 2.3 31.8 ± 2.7 4.8 ± 0.9 7.1 ± 1.0 9.4 ± 1.2 0.43 ± 0.05 0.17 ± 0.04 0.09 ± 0.04
9.5 ± 2.2 12.2 ± 3.6 21.3 ± 5.7 3.8 ± 1.0 5.1 ± 1.2 7.0 ± 1.6 0.40 ± 0.11 0.24 ± 0.06 0.08 ± 0.09
8.7 ± 1.017.2 ± 4.6 2.9 ± 0.5 4.4 ± 0.9 5.5 ± 1.3 0.47 ± 0.12 0.23 ± 0.11 0.06 ± 0.09
Control
Regenerate
Control
13.0 ± 2.5 16.2 ± 2.7 33.2 ± 2.6 4.9 ± 1.0 7.6 ± 1.3 9.9 ± 1.0 0.44 ± 0.05 0.17 ± 0.04 0.08 ± 0.03
Figure 2. Box plot of cortical bone density in kg/m3: (A) between paired regenerate and control specimens in the two anatomical
positions; (B) between paired regenerate and control specimens in two relative positions within the mandible; (C) between
regenerate and control cortical bone, irrespective of position
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term

Page 5

Mechanical properties of canine cortical bone
Figure 3. Principal stiffness direction distribution through the cortical samples is represented by Rose diagrams: (A) principal
direction of maximum stiffness on the buccal and lingual sides; (B) principal direction of maximum stiffness through the 35 cortical
specimens
Figure 4. Box plot of elastic moduli in GPa: (A) between anatomical positions in the regenerate cortical bone; (B) between
regenerate and control cortical bone specimens; (C) averaged through the 35 cylindrical cortical bone specimens
Figure 5. Box plot of shear moduli in GPa: (A) between anatomical positions in the regenerate cortical bone specimens; (B) between
regenerate and control cortical bone specimens; (C) averaged among 35 cortical bone specimens
3.4. Shear moduli
No significant differences in shear moduli were found
in either G31 (F = 2.03, p = 0.115) or G23 (F = 1.036,
p = 0.405) among the five dogs. In contrast, G12 in
dog 1 was consistently different to that in dogs 2, 3,
4 and 5 (F = 5.922, p = 0.001). In addition, significant
differences in shear moduli G12 (p = 0.048), G31 (p =
0.037) and G23 (p = 0.028) were found between the
buccal and lingual anatomical positions in the regenerate
cortical bone (Figure 5A). No significant differences
were found in shear moduli between the buccal and
lingual anatomical positions in the control cortical
bone (Table 2). Furthermore, no significant differences
in the regenerate cortical bone between the alveolar
and basal relative positions were found in G12 or G23
(Table 2); however, statistical differences were found in
G31 (p = 0.035). In contrast, a significant difference in
G23(p = 0.008) between alveolar and basal positions was
the only difference found for the control cortical bone.
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term

Page 6

U. Zapata et al.
Figure 6. Box plot of Poisson’s ratios: (A) υ12 among subjects; (B) between regenerate and control cortical bone specimens;
(C) through 35 cylindrical cortical bone specimens
Complementarily, significant differences in shear moduli
G12 (p = 0.003), G31 (p = 0.001) and G23 (p = 0.001)
were found between the regenerate cortical bone and
the control cortical bone (Table 2, Figure 5B). Statistical
differences among G12, G23andG31(F = 63.9,p = 0.001)
were found (Figure 5C).
3.5. Poisson’s ratio
No statistical differences in Poisson’s ratios υ13(F = 2.6,
p = 0.057) and υ23 (F = 2.2, p = 0.091) were found
among the five dogs. However, a significant difference
for υ12 (F = 4.3, p = 0.007) was found in dog 1 when
compared with the same characteristic in dogs 4 and 5
(Figure 6A). In the same way, no significant differences
in Poisson’s ratios υ12(p = 0.504), υ23(p = 0.336) and
υ31 (p = 0.611) were found in the regenerate cortical
bone between paired buccal and lingual anatomical
positions. However, no significant differences were found
between the buccal and lingual sides in the control
cortical bone for Poisson’s ratio υ13 (p = 0.858) and
υ23 (p = 0.096), although a significant difference was
reported in υ31(p = 0.018) (Table 2). Correspondingly,
a test between relative positions in the regenerate cortical
bone showed no significant differences in any of the
Poisson’s ratios υ12(p = 0.193), υ23(p = 0.052) and υ31
(p = 0.747). Complementarily, no significant differences
in the control cortical bone between relative positions
were found for either υ12(p = 0.123) or υ23(p = 0.208),
although a significant difference was reported for υ31
(p = 0.020). Finally, no differences in Poisson’s ratio
for either υ12(F = 1.54, p = 0.224), or υ23(F = 1.299,
p = 0.263) were found between the regenerate and
control cortical bone specimens (Figure 6B), although
significant differences were found for υ13 (F = 6.143,
p = 0.018). Statistical differences were present among all
three principal planes (Figure 6C).
4. Discussion
Discrepancies in sample size between control and
regenerate sides were due to failure in obtaining
cylindrical regenerate cortical samples from the buccal
side because of either the presence of soft tissue inclusions
or low cortical thickness. However, the specimen number
was statistically representative to allow comparisons
among groups.
4.1. Differences among anatomical positions
Although the MBT device is placed in the buccal side,
our results showed that the mechanical integrity of the
regenerate cortical bone was lower on the buccal than the
lingual side of the mandible. This finding could be due to
theeffect ofthree factorsthatcanacteither independently
or jointly: first, the device could produce a stress-shielding
effect of the regenerate cortical bone on the buccal side,
dampening potential mechanical stimulus; this condition
should improve after the device is removed (Kennady
et al., 1989; Laftman et al., 1989). Second, the relative
position of the device within the mouth could affect the
mechanicalqualityoftheregeneratecorticalbonebecause
of its effect on inflammatory processes due to foreign body
reaction after device placement (Anderson, 1988). Third,
the screws attached to the transport disc on the facial side
may affect the quality of the new cortical bone generated
by traction on that side.
Differences in elastic mechanical properties between
relative positions, alveolar and basal, were present mainly
amongthecontrolspecimensratherthanintheregenerate
cortical bone. This finding in the control bone could be
due to bone adaptations to normal functional patterns
of mandibular loading, inertia at the cross section, and
the average direction of flexion load, which produces
compressive stresses at the basal cortical bone and
tensile stresses at the alveolar cortical bone (Alvarez-
Arenal et al., 2009). Alternately, lower chewing forces
on the regenerate side, due to the animal favouring the
unoperated side, and the stress-shielding effect from the
reconstruction plate could also result in overall reduced
cortical bone formation in the regenerate.
4.2. Differences between regenerate and control
cortical bone
Although very little is known about the biomechanical
characteristics of regenerate cortical bone after MBT,
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term

Page 7

Mechanical properties of canine cortical bone
it
bone
properties, microstructure and morphological features
of the original cortical bone at the end of the
mandibular distraction process (Schickendantz et al.,
1992). Accordingly, several studies have tried to assess
differences between regenerate and control cortical
bone after distraction osteogensis by performing global
biomechanical tests that ignore the anisotropic behaviour
of the cortical bone and consequently assigning a global
strength factor to the regenerate cortical bone (Ayoub
et al., 2001; Li et al., 2002; Gomez et al., 2005; Kunz
et al., 2006). Although, in the present study, the elastic
mechanical properties of the control cortical bone were
higher than the regenerate cortical bone, these differences
varied depending on which mechanical property was
compared. For instance, density in the regenerate cortical
bone was 88.44% of the density in the control cortical
bone; elastic modulus E3in the regenerated cortical bone
was 73.57% of the stiffness in the control cortical bone;
and shear modulus G12 in the regenerate cortical bone
was 77.08% of the control cortical bone.
has been
tends
assumed
tocompletely
that the
restore
distracted
the
cortical
mechanical
4.3. Mechanical properties and anisotropy
Although it has been reported that canine mandibular
cortical bone is elastically isotropic (Ashman et al., 1985),
several studies have reported anisotropic behaviour in
mammalian mandibles of several species (Ashman et al.,
1987; Schwartz-Dabney and Dechow, 2002), and in
canine long bones (Ashman et al., 1984). The present
study quantified anisotropy using three complementary
estimates: first, by statistically comparing the three
principal elastic moduli; second, by comparing relative
stiffness between axes by using the ratio E2:E3 to
measure the relative bone anisotropy in the cortical plane;
and third, by comparing the directions of maximum
stiffness between regenerate and control cortical bone.
The principal elastic moduli E1and E2were statistically
different in the control cortical bone (F = 19.1, p <
0.05), suggesting an orthotropic organization at a
tissue level, whereas they were statistically similar
in the regenerate cortical bone (F = 2.5; p = 0.13),
implying a transversal isotropic organization (Table 3).
The relative bone anisotropy in the cortical bone E2:E3
was calculated as 0.52 ± 0.12 and 0.50 ± 0.06 for
the regenerate and control cortical bone, respectively,
demonstrating similar differences in these orientations.
Lastly, the direction of maximum stiffness in the
regenerate and the control cortical bone was found
to have the same principal orientation with respect to
the lower border of the mandibular corpus. Overall, the
calculated values for the principal elastic moduli of the
control and regenerate cortical bone were numerically
different in all directions, but the control cortical bone
showed orthotropic characteristics, while the regenerate
cortical bone showed transversely isotropic characteristics
because of the similarity between elastic moduli E1and
E2.
4.4. Relationship between density
and mechanical properties
The mechanical properties of cortical bone are affected
by the density, grade of mineralization and matrix
organization (Rho et al., 1998). In general, bone density
increases with time as a result of both remodelling and
mineralization,whichresultsinanincreaseinthestrength
and stiffness of the cortical regenerate bone (Walsh
et al., 1994). Increases in bone density are an important
part of the bone healing process following MBT (Harp
et al., 1994; Kunz et al., 2006). Density is lower in the
regenerate cortical bone at the end of the consolidation
time than the control cortical bone (Kunz et al., 2006).
Our results showed that average density was about 11%
lower in the regenerate cortical bone than in the control
cortical bone after 12 weeks of consolidation, which has
density values similar to those reported previously for
human mandibular cortical bone (Schwartz-Dabney and
Dechow, 2003).
Although elastic moduli were deduced using mate-
rial density characteristics, different relationship patterns
Table 3. Canine elastic mechanical properties for both regenerate and control cortical bone compared with averaged mechanical
properties reported previously
Mechanical
property
Regenerate
cortical bone
Control
cortical bone
Canine
femura
Canine
mandibleb
Human
mandiblec
Human
mandibled
E1(GPa)
E2(GPa)
E3(GPa)
G12(GPa)
G31(GPa)
G23(GPa)
υ12
υ13
υ23
9.9 ± 2.7
11.7 ± 3.6
23.1 ± 6.9
3.7 ± 1.1
5.6 ± 1.4
7.0 ± 1.7
0.45 ± 0.09
0.20 ± 0.07
0.08 ± 0.08
12.6 ± 2.2
15.7 ± 2.4
31.4 ± 3.1
4.8 ± 0.9
7.2 ± 1.1
9.2 ± 1.3
0.42 ± 0.05
0.16 ± 0.03
0.10 ± 0.04
12.8
15.6
20.1
4.68
5.68
6.67
0.282
0.289
0.265
7.39
7.39
7.39
2.63
2.63
2.63
0.403
0.403
0.403
10.8
13.3
19.4
3.81
4.12
4.63
0.309
0.249
0.224
12.7
17.9
22.8
5.0
5.5
7.4
0.18
0.31
0.28
aAshman et al., 1984.
bAshman et al., 1985.
cAshman and Van Buskirk, 1987.
dSchwartz-Dabney and Dechow, 2003.
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term

Page 8

U. Zapata et al.
Figure 7. Scatterplot of density in kg/m3and principal elastic moduli in MPa: (A) at the regenerate cortical bone; (B) at the control
cortical bone
were presented between control and regenerated spec-
imens. In the control cortical bone, the trends for the
minimum, middle and maximum elastic moduli when
compared with bone density are parallel (Figure 7B). As
the bone density increases, so do the values for E1, E2
and E3. In addition, E1and E2have similar tendencies,
supporting their orthotropic characteristics. These results
have the same tendency as the relationship reported
by Schaffler and Burr (1988). They reported a positive
correlation between elastic moduli obtained from ten-
sion tests and density in bovine cortical bone samples,
though density was calculated from measured porosity
instead of being measured directly. On the contrary,
for the regenerate cortical bone, although the values
for E1 and E2 are close, their tendencies are conver-
gent (Figure 7A). This geometrical characteristic seems
to suggest an immature tissue representation. In addi-
tion, it is possible to observe two groups of data for
every tendency line that may correspond to the previ-
ous described difference in density between anatomical
positions.
5. Conclusions
As far as it is possible to determine, this is the first
report to measure the elastic constants of cortical bone
regenerate after MBT osteogenesis by using the non-
destructive ultrasonic technique. The results suggest that
the control mandibular cortical bone is a transverse
isotropic material, whereas the regenerate cortical bone is
transversally isotropic with structural elements that tend
to be orientated preferentially in the same direction as
the distraction process, providing maximum stiffness in
that direction. Although there are statistical differences
in elastic moduli between regenerate and control
cortical bone, both tissues remain anisotropic and the
numerical differences must be due to different grades of
mineralization.
Future studies relating to the biomechanics of the
regenerate tissues after the MBT process must consider
correspondingly the biomechanical differences in the
trabecular bone, the relationship between microstructure
and bony mechanical properties, patterns of blood supply
in the regenerate tissues and their relationship with
mechanical properties, and the effect of growth factors
on the microstructure and mechanical properties of the
new regenerate tissues.
Acknowledgements
This work was supported by the NIH/NIDCR (Grant Nos R42
DE015437-03 and R43 DE017259-05) and a fellowship from
the Departamento Administrativo de Ciencia, Tecnolog´ ıa e
Innovaci´ on (COLCIENCIAS), Colombia.
References
Alvarez-Arenal A, Lasheras FS, Fern´ andez
EM, et al. 2009; A jaw model for the
study of the mandibular flexure taking
into account the anisotropy of the bone.
Math Comput Modelling 50: 695–704.
Anderson JM. 1988; Inflammatory response
to implants. ASAIO Trans 34: 101–107.
Ashman RB, Corin JD, Turner CH. 1987;
Elastic properties of cancellous bone:
measurement by an ultrasonic technique.
J Biomech 20: 979–986.
Ashman RB, Cowin SC, Van Buskirk WC,
et al. 1984; A continuous wave technique
forthemeasurement
properties of cortical bone. J Biomech 17:
349–361.
Ashman RB, Rosinia G, Cowin SC. 1985;
The bone tissue of the canine mandible
is elastically isotropic. J Biomech 18:
717–721.
oftheelastic
Ashman RB, Van Buskirk WC. 1987; The
elastic properties of a human mandible.
Adv Dental Res 1: 64–67.
Ayoub AF, Richardson W, Koppel D, et al.
2001; Segmental mandibular reconstruc-
tion by microincremental automatic dis-
traction osteogenesis: an animal study. Br
J Oral Maxillofac Surg 39: 356–364.
Black J, Perdigon P, Brown N, et al. 1984;
Stiffness and strength of fracture callus.
Relative rates of mechanical maturation
as evaluated by a uniaxial tensile test. Clin
Orthop Rel Res 182: 278–288.
Casta˜ no MC, Zapata U, Pedroza A, et al.
2002; Creation of a three-dimensional
modelof the mandible and the TMJ in vivo
by means of the finite element method. Int
J Comput Dentist 5: 87–99.
Cho BC, Moon JH, Chung HY, et al. 2003;
The bone regenerative effect of growth
hormone on consolidation in mandibular
distraction osteogenesis of a dog model.
J Craniofac Surg 14: 417–425.
Claes LE, Cunningham JL. 2009; Monitoring
the mechanical properties of healing bone.
Clin Orthop Rel Res 467: 1964–1971.
Costantino PD, Friedman CD, Shindo ML,
et al.1993; Experimental
regrowth bydistraction
long-term results. Arch Otolaryngol Head
Neck Surg 119: 511–516.
de Zee M, Dalstra M, Cattaneo PM, et al.
2007; Validation of a musculo-skeletal
model of the mandible and its application
to mandibular distraction osteogenesis.
J Biomech 40: 1192–1201.
Degidi M, Pieri F, Marchetti C, et al. 2004;
Immediate loading of dental implants
placedindistractedbone:acasereport.Int
J Oral Maxillofac Implants 19: 448–454.
mandibular
osteogenesis:
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term

Page 9

Mechanical properties of canine cortical bone
Fitch RD,
1996;
distraction technique on bone and muscle
in a canine model: a preliminary report.
Iowa Orthop J 16: 10–19.
Gantous A,Phillips JH,Catton P,et al.1994;
Distraction osteogenesis in the irradiated
canine mandible. Plastic Reconstruct Surg
93: 164–168.
Gomez DF, Sant’Anna EF, Leven RM, et al.
2005;Microstructural
evaluation of regenerate tissue during
the consolidation period after vertical
mandibular ramus distraction. J Craniofac
Surg 16: 805–811.
Harp JH, Aronson J,
Noninvasive determination
stiffness for distraction osteogenesis by
quantitative computed tomography scans.
Clin Orthop Rel Res 301: 42–48.
Holmes SB, Lloyd T,
2002; Distraction osteogenesis of the
mandible in the previously irradiated
patient.
J OralMaxillofac
305–309.
Ilizarov GA. 1989a; The tension-stress effect
on the genesis and growth of tissues. Part
I. The influence of stability of fixation and
soft-tissue preservation. Clin Orthop Rel
Res 238: 249–281.
Ilizarov GA. 1989b;
effectonthegenesis andgrowth oftissues:
Part II. The influence of the rate and
frequency of distraction. Clin Orthop Rel
Res 239: 263–285.
Imola MJ, Hamlar DD, Thatcher G, et al.
2002;The versatility
osteogenesis in craniofacial surgery. Arch
Facial Plastic Surg 4: 8–19.
Karp NS, Thorne CH, McCarthy JG, et al.
1990; Bonelenghthening
craniofacial skeleton. Ann Plastic Surg 24:
231–237.
Kennady MC, Tucker MR, Lester GE, et al.
1989; Histomorphometric evaluation of
stress shielding in mandibular continuity
defects treated with rigid fixation plates
and bone grafts. Int J Oral Maxillofac Surg
18: 170–174.
Thompson JG,
Theeffects
Rizk WS,
of the
et al.
Ilizarov
andstrength
Hollis M.1994;
bone of
Coghlan KM,
et al.
Surg
60:
Thetension–stress
ofdistraction
in the
Kunz C, Adolphs N, B¨ uscher P, et al. 2006;
Mineralizationand
pertiesof the
distraction wound
molding. Int J Oral Maxillofac Surg 35:
822–827.
Kuriakose MA, Shnayder Y, DeLacure MD.
2003;Reconstruction
mandibular defects by distraction osteoge-
nesis for mandibular reconstruction. Head
Neck 25: 816–824.
Laftman P, Nilsson OS, Brosjo O,et al.1989;
Stress shielding by rigid fixation studied
in osteotomized rabbit tibiae. Acta Orthop
Scand 60: 718–722.
Li J, Hu J, Wang D, et al. 2002; Biomechan-
ical properties of regenerated bone by
mandibular distraction osteogenesis. Chin
J Traumatol (Engl) 5: 67–70.
Lipson SF, Katz JL. 1984; The relationship
between elastic properties and microstruc-
ture of bovine cortical bone. J Biomech 17:
231–240.
Marinescu R, Daegling DJ, Rapoff AJ. 2005;
Finite-element modelingoftheanthropoid
mandible:the
boundary conditions. Anat Rec A 283:
300–309.
Orbay JL, Frankel VH, Finkle JE, et al. 1992;
Canine leg lengthening by the Ilizarov
technique: a biomechanical, radiologic,
and morphologic study. Clin Orthop Rel
Res 278: 265–273.
Perrott DH, Rahn B, Wahl D, et al. 2003;
Development of a mechanical testing
systemfora mandibular
wound. Int J Oral Maxillofac Surg 32:
523–527.
Reilly DT, Burstein AH. 1975; The elastic
and ultimate properties of compact bone
tissue. J Biomech 8: 393–405.
Rho JY. 1996; An ultrasonic method for
measuring the elastic properties of human
tibialcorticaland
Ultrasonics 34: 777–783.
Rho JY, Kuhn-Spearing L, Zioupos P. 1998;
Mechanical properties and the hierarchi-
cal structure of bone. Med Eng Physics 20:
92–102.
mechanical
canine
following
pro-
mandible
acute
ofsegmental
effectsof altered
distraction
cancellousbone.
Robertson DM, Smith DC. 1978; Compres-
sive strength of mandibular bone as a
function of microstructure and strain rate.
J Biomech 11: 455–471.
Rubio-Bueno P, Padr´ on A,
2000; Distraction
the ascending ramus for mandibular
hypoplasia using extraoral or intraoral
devices: a report of 8 cases. J Oral
Maxillofac Surg 58: 593–599.
Sacco AG,Chepeha DB.2007;Currentstatus
of transport-disc-distraction osteogenesis
for mandibular reconstruction. Lancet
Oncol 8: 323–330.
Schaffler MB, Burr DB. 1988; Stiffness of
compact bone: effects of porosity and
density. J Biomech 21: 13–16.
Schickendantz MS, Watson JT,
et al. 1992; A model for evaluating
the strength of bones lengthened by
distraction osteogenesis. Clin Orthop Rel
Res 275: 248–252.
Schwartz-Dabney CL,
Edentulation alters material properties of
cortical bone in the human mandible.
J Dent Res 81: 613–617.
Schwartz-Dabney CL,
Variations in cortical
perties throughout the human dentate
mandible. Am J Phys Anthropol 120:
252–277.
Walsh WR, Hamdy RC,
1994; Biomechanical and physical pro-
perties of lengthened bone in a canine
model.
ClinOrthop
230–238.
Zioupus PSC, An YH (eds). 2000; Factors
Affecting Mechanical Properties of Bone.
CRC: Boca Raton, FL.
Zysset PK, Edward Guo X, Edward Hoffler C,
et al.1999; Elastic
hardness of cortical and trabecular bone
lamellae measured by nanoindentation
in the human femur. J Biomech 32:
1005–1012.
Villa E,
et al.
osteogenesisof
Sferra JJ,
Dechow PC. 2002;
Dechow PC.
material
2003;
pro-
Ehrlich MG.
Rel Res
306:
modulus and
Copyright  2011 John Wiley & Sons, Ltd.
J Tissue Eng Regen Med (2011).
DOI: 10.1002/term