Full text of DOI:10.1359/jbmr.2001.16.4.671

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 16, Number 4, 2001
© 2001 American Society for Bone and Mineral Research
Low-Intensity Pulsed Ultrasound Accelerates Rat Femoral
Fracture Healing by Acting on the Various Cellular
Reactions in the Fracture Callus
YOSHIAKI AZUMA,1 MASAYA ITO,1 YOSHIFUMI HARADA,1 HIDEKO TAKAGI,1
TOMOHIRO OHTA,1 and SEIYA JINGUSHI2
ABSTRACT
Low-intensity pulsed ultrasound (LIPUS) has been shown to accelerate fracture healing in both animal models
and clinical trials, but the mechanism of action remains unclear. In fracture healing, various consecutive
cellular reactions occurred until repair. We investigated whether the advanced effects of LIPUS depended on
the duration and timing of LIPUS treatment in a rat closed femoral fracture model to determine the target
of LIPUS in the healing process. Sixty-nine Long-Evans male rats that have bilateral closed femoral fractures
were used. The right femur was exposed to LIPUS (30 mW/cm2 spatial and temporal average [SATA], for 20
minutes/day), and the left femur was used as a control. Rats were divided into four groups according to timing
and duration of treatment (Ph-1, days 1–8; Ph-2, days 9–16; Ph-3, days 17–24; throughout [T], days 1–24 after
the fracture). Animals were killed on day 25. After radiographs and microfocus X-ray computed tomography
(␮CT) tomograms were taken, the hard callus area (HCA), bone mineral content (BMC) at the fracture site,
and mechanical torsion properties were measured, and histological analysis was conducted. Interestingly, the
maximum torque of the LIPUS-treated femur was significantly greater than that of the controls in all groups
without any changes in HCA and BMC. The multiviewing of three-dimensional (3D) ␮CT reconstructions and
histology supported our findings that the partial LIPUS treatment time was able to accelerate healing, but
longer treatment was more effective. These results suggest that LIPUS acts on some cellular reactions involved
in each phase of the healing process such as inflammatory reaction, angiogenesis, chondrogenesis, intramem-
branous ossification, endochondral ossification, and bone remodeling. (J Bone Miner Res 2001;16:671–680)
Key words: low-intensity pulsed ultrasound, fracture healing, fracture callus, mechanical properties, me-
chanical stress
agnostic ultrasound employs much lower intensities (1–50
mW/cm2), which are chosen specifically to avoid tissue
heating.(1)
Fracture healing has been accelerated by exposure to a
specific pulse of extremely low-intensity pulsed ultrasound
INTRODUCTION
LTRASOUND IS used widely in surgery, therapeutics, and
diagnostics.(1) The frequency used ranges from 0.8 to
15.0 MHz, which is based on considerations of sound ab-
U
sorption, penetration, and resolution. The intensities are (LIPUS) in the latter power range in both animal fracture
high for surgical and therapeutic applications (1–50 W/cm2) models and clinical trials.(3–9) For example, Heckman et al.
and are designed to cause significant tissue heating.(2) Di- reported that LIPUS treatment for 20 minutes/day at 30
1Teijin Institute for Biomedical Research, Teijin Limited, Tokyo, Japan.
2Department of Orthopedic Surgery, Graduate School of Medical Science, Kyushu University, Fukuoka, Japan.
671

672
AZUMA ET AL.
mW/cm2 led to a significant 24% reduction in the time
required for clinical healing in a prospective randomized
placebo-controlled study of closed or grade I open tibial
fractures.(8) Additionally, Kristiansen et al. showed that the
time until union was 38% shorter for patients treated with
LIPUS than for those treated with placebo in a multicenter
prospective randomized placebo-controlled clinical trial of
dorsally angulated fractures of the distal radius.(9) However,
the physical process through which LIPUS stimulates living
tissue remains unclear. LIPUS is a form of mechanical
energy that can be transmitted into living tissue as high-
frequency acoustical pressure waves. The micromechanical
strains produced by these pressure waves in living tissue can
result in biochemical events at the cellular level.(10–12) Du-
arte indicated that a pulsed ultrasound intensity of 49.6
mW/cm2 and 57 mW/cm2 spatial average and temporal
average (SATA) was nonthermal.(4) In this experiment, we
used LIPUS (30 mW/cm2), which is considered to have
little thermal effect and to produce stable cavitation and
streaming. Wang et al. reported a technique for exposing
standard closed femoral fractures to LIPUS and confirmed
that LIPUS stimulation enhanced the mechanical properties
of the healing callus.(7) Their observations imply that cells
in the fracture callus sense and respond to the mechanical
energy transferred by the LIPUS.
Fracture healing occurs in a stepwise process of various
cellular reactions until the fracture has been repaired.(13)
Cell types, such as endothelial cells, fibroblasts, osteoblasts,
chondrocytes, and osteocytes, all of which respond to me-
chanical stress or mechanical strain,(14–18) play important
roles in the fracture healing process.(19) The mechanical
energy transferred by LIPUS therefore might act on various
processes such as inflammatory reaction, angiogenesis,
chondrogenesis, intramembranous ossification, endochon-
dral ossification, or bone remodeling in the fracture callus.
In previous experiments using a rat closed femoral frac-
ture model, we found accelerating effects of daily LIPUS
treatment until 21 days after the fracture (in preparation).
Because in this model a series of stage-specific cellular
reactions were observed until repair,(20) we investigated
whether the advanced effects of LIPUS depended on dura-
tion and timing of LIPUS treatment in this model to deter-
mine the target reaction of LIPUS in the fracture healing
process in vivo. In the present study, we divided the interval
up to 25 days after the fracture into three periods (Ph-1,
Ph-2, and Ph-3), each of which involved different essential
reactions for fracture healing,(20) and investigated the ef-
fects of partial treatment with LIPUS on the fracture heal-
ing. If the advanced effects of LIPUS on fracture healing are
based on an effect on a specific reaction, the efficiency of
acceleration should depend on the timing of the partial
LIPUS treatment. In this study, we investigated the rela-
tionship between the timing of the partial LIPUS treatment
and the efficiency of accelerating action by estimating me-
chanical properties, bone mass, histology, and the micro-
structure of the callus, the latter by using the three-
dimensional (3D) microfocus X-ray computed tomography
(␮CT) technique.
FIG. 1. Experimental schedule for LIPUS treatment regimens. The
animals in each group were treated with LIPUS as follows: the duration
was in the Ph-1 group from day 1 to 8 after the fracture (8 days), in the
Ph-2 group from day 9 to 16 after the fracture (8 days), in the Ph-3
group from day 17 to 24 after the fracture (8 days), and in the T group
from day 1 to 24 after the fracture (24 days). All remaining animals
were killed on day 25 after the fracture.
MATERIALS AND METHODS
Closed femoral fracture model
A total of 69 skeletally mature 10-week-old male Long-
Evans rats (Charles-River Laboratories, Inc., Wilmington,
MA, USA), weighing 300–350 g, were used for this study.
All rats were housed at 24 Ϯ 2°C, a relative humidity of
55 Ϯ 10% with a 12-h/12-h light/dark cycle. The rats were
fed standard rat chow (CE2) obtained from Clea Japan, Inc.
(Tokyo, Japan) and were given water ad libitum. The closed
femoral fracture models were produced by the method pre-
viously described.(21) Under anesthesia induced by intra-
peritoneal (ip) injection with sodium pentobarbital (50
mg/kg per ml; Dinabot, Inc., Osaka, Japan), a medial
parapatellar arthrotomy was made, and the patella was then
dislocated laterally. The medullary canal was entered
through the intercondylar notch and reamed with a 19G
needle. A 1 mm ϫ 28 mm Kirschner wire (Natsume Co.,
Ltd., Tokyo, Japan) was then inserted into the canal. The
patella was relocated, and then the soft tissues were closed
with no. 4.0 silk sutures. The stabilized femur was placed in
a three-point bending device (Arizono Machine Co., Ltd.,
Kagoshima, Japan), and the femur was fractured by the
force of a 500-g weight dropped from a height of 35 cm.
Bilateral fractures were made in all animals. Radiographs
were used to confirm of the position and orientation of each
fracture. Because we used rats that satisfied the inclusion
criteria only, the sample size in each group was unequal.
Three rats in the Ph-1 group and one in the throughout (T)
group did not satisfy the criteria. Because those rats were
not used for LIPUS treatment, a total of 65 animals were
used as shown in Fig. 1. Our Institutional Animal Care and
Use Committee approved all procedures.
LIPUS signals
The LIPUS exposure system used, which was a modifi-
cation of the device used in experiments by Wang et al.,(7)
was made by Medical Engineering Research Laboratories of

ULTRASOUND ACCELERATES RAT FEMORAL FRACTURE REPAIR
673
Teijin, Ltd. (Tokyo, Japan). It consisted of a function gen- of the fracture line, which was measured with software for
erator (model HP33120A; Hewlett-Packard Japan, Ltd., To- high-resolution analysis. The CV for the measurement of
kyo, Japan), a power supply, and six lead-zirconate titanate the BMC of standard samples by this technique was 0.8%.
transducers (PZT-4, effective area of 3.88 cm2). With it, six Specimens were stored in saline-soaked gauze at Ϫ20°C.
animals could be exposed to LIPUS at the same time. The
LIPUS signal was generated with a transducer and was
composed of a 200-␮s burst sine wave of 1.5 MHz repeating
Hard callus area
at a frequency of 1.0 kHz. The 200-␮s burst of pressure
Radiographic signals of each femur and the standard scale
pulses was followed by an off time of 800 ␮s and was were captured as digitized images with a scanner and stored
repeated every millisecond. The LIPUS intensity was 30 on a Macintosh computer using Adobe PhotoShop version 4
mW/cm2 as an SATA. The LIPUS signal and times per day software. The area of the outer hard callus was measured by
used in this study were the same as those recommended for tracing the callus portion with “National Institutes of Health
a clinical device (Sonic Accelerated Fracture Healing Sys- (NIH) image version 1.6” image analysis software (NIH,
tem [SAFHS; Exogen, Inc., Piscataway, NJ, USA]). The Bethesda, MD, USA). The area represents the number of
LIPUS signal conditions were calibrated with a hydrophone pixels within each tracing. The standard scale also was
system that consisted of an oscilloscope (model OS-111; measured, and the pixel value of each hard callus area
Hewlett-Packard Japan, Ltd.) and a PZT-4 transducer for (HCA) was transferred to the value of the real scale.
detection. The LIPUS intensity was calibrated with an ul-
trasound power meter (model 101; OR Instruments, Inc.,
Seattle, WA, USA) before LIPUS exposure to the animals.
Mechanical testing
The interexperiment CVs of each transducer was within
10% for the LIPUS intensity.
Just before mechanical testing, the specimens were
thawed at room temperature for 2 h. Both bone ends were
embedded in polyacrylic resins (GC-OSTORON; GC Den-
tal Products Co., Ltd., Aichi, Japan). A custom-made jig
ensured consistent alignment of the bone axis with the axis
of the testing machine. All specimens were tested to failure
in torsion at room temperature on an electromechanical
testing machine (model MZ-500D; Maruto Machine, Inc.,
Tokyo, Japan) at the rate of 1.5°/s with 111.5 g of weight as
axial load. Maximal torque until failure and torsional stiff-
ness (the tangent at the point of maximal slope) were
calculated from the load-deformation curve. The biome-
chanical stages of fracture union were determined based on
pattern of failure according to White et al.(22)
Duration and timing of LIPUS treatment
The animals were treated with LIPUS while they were
under ip anesthesia with ketamine (20–60 mg/kg) and xy-
lazine (2.5 mg/kg). The anesthetized rats were placed on
their back on the holder. Ultrasound gel (Nippon Kohden
Co., Ltd., Tokyo, Japan) was applied to the shaved fracture
area in the medial aspects of both thighs of all animals. The
transducer was placed close to the skin covered with ultra-
sound gel at the femoral fracture site. The animals were
exposed to LIPUS once daily for 20 minutes. The right
femur of the rats was exposed to the LIPUS transducer, and
the left femur was used as the nontreated contralateral
control. The duration and timings of the LIPUS treatment
are shown in Fig. 1. The animals in each group were treated
with LIPUS as follows: LIPUS treatments were performed
in the Ph-1 group for 8 days, from day 1 to 8 after fracture;
in the Ph-2 group for 8 days, from day 9 to 16 after fracture;
in the Ph-3 group for 8 days, from day 17 to 24 after
fracture; and in the T group for 24 days, from day 1 to 24
after fracture. Animals were killed on day 25 after the
fracture.
Histological evaluation
Three rats per each group were killed, and the femora
were collected on day 25 after the fracture for histological
evaluation. To observe the effect of LIPUS in progress on
day 9 and day 17 after fracture, we killed three rats in the
Ph-1 and Ph-2 groups, respectively. The femora were fixed
for several days in 10% neutral buffered formalin. After the
␮CT images had been obtained, the femora were treated
with 10% formic acid for 1 h, transiently. Then, the femora
were decalcified with 10% EDTA solution and embedded in
paraffin. Three to five sections (6 ␮m thick) were cut
through the long axis of each femur in the sagittal plane and
stained with Masson’s trichrome stain and hematoxylin and
Radiography and bone mineral content at the
fracture site
The femora were collected on day 25 after the fracture. eosin stain. The section that had the maximum diameter was
All femora were radiographed with a soft X-ray system chosen as the representative section for each fracture.
(model CSM; Softex Co., Ltd., Tokyo, Japan). After the soft
tissues had been dissected from the femora, the intramed-
ullary pins were removed. Bone mineral content (BMC) in
3D ␮CT analysis
the area near the fracture site was measured with a dual-
The ␮CT (model NX-CP-C80H-IL) equipment that pro-
energy X-ray absorptiometer (DXA; model QDR-2000; Ho- duced the tomograms had been especially developed by
logic, Inc., Waltham, MA, USA). The scan field size was Nittetsu ELEX Co., Ltd. (Osaka, Japan). The device has a
3.48 cm2 ϫ 1.41 cm2, and the resolution was 0.0254 ϫ maximum experimental spatial resolution of 2.5 ␮m and a
0.0127 cm2. The area near the fracture site of each femur minimum slice thickness of 5 ␮m for measurements of a
was determined as the area extending 3.6 mm on either side ceramic standard sample. The fan beam of ␮CT has a small

674
AZUMA ET AL.
focal spot size, 3 ␮m in diameter, a conventional microfo-
cus open vacuum-type X-ray source, rotating sample stage,
1024 linear response image sensors, and image intensifier
(image sensors). A total of 100 cross-sectional tomograms
per femur were obtained with a slice thickness of 80 ␮m and
reconstructed at 512 ϫ 512 pixels, by use of an acceleration
voltage of 30 kV and current of 0.1 mA. To evaluate the
bone bridging at the fracture site by use of a 3D reconstruc-
tion technique, we took 50 slices extending (4000 ␮m) on
either side of the fracture line. The 3D reconstruction of 100
tomograms was performed by the volume-rendering method
(software, VIP-Station; Teijin System Technology, Inc.,
Kanagawa, Japan) using
a computer (model SUN
SPARK-5; Sun Microsystems, Inc., Palo Alto, CA, USA).
By use of the computer software, the threshold of each
image was kept at the same level to observe trabecular bone
and cortex of the healing callus. At this threshold, we never
detected the soft callus of the femur. Electronic sections
were cut through the long axis of each femur in the sagittal
plane on 3D reconstructed images. The home sagittal plane
(0°) was identified as that which was used to cut through the
long axis of the original 3D reconstructed image of each
femur. To estimate the bone bridging in detail, we rotated
the sagittal plane clockwise about the long axis at 60°
increments of pitch (0°, 60°, 120°, 180°, 240°, and 300°)
from the home position in a circuit and cut electronic
sections through the long axis of each femur in each sagittal
plane.
Statistical analysis
The primary analyses conducted in this study were in-
traexperimental group comparisons of the treated fractures
with the untreated contralateral fractures. Paired t-testing
was used for analyses of HCA, BMC, and mechanical
measurement data. The significance of interexperimental
group comparison of mechanical measurement data was
FIG. 2. Effects of LIPUS treatment of various durations and timings
on (A) torsional torque and (B) torsional stiffness of fracture callus on
day 25 in rat femoral fractures. The torsional torque of LIPUS-treated
femurs (closed bars) was significantly higher than that of the nontreated
femurs (open bars) in the T group (n ϭ 11), Ph-1 group (n ϭ 9), Ph-2
group (n ϭ 12), and Ph-3 group (n ϭ 12). The horizontal bars represent
determined by a one-way analysis of variance (ANOVA) Ϯ 1 SD. The value of p Ͻ 0.01 for each treatment group compared with
the control side. Paired t-testing was used for statistical analysis to
compare between the LIPUS-treated side and -nontreated side in each
group. The T group showed a significant difference as compared with
the partially treated groups. One-way ANOVA was used for statistical
test. A confidence level of 95% (p Ͻ 0.05) was chosen for
significance.
analysis to test for significance versus the T group.
RESULTS
Mechanical properties of fracture callus
The maximal torque and stiffness in torsion of the frac-
tured femur on the LIPUS-treated side was significantly
Radiographs and 3D ␮CT analysis
higher than that of the contralateral control side in all groups
As shown in Fig. 3, in the LIPUS-treated femur, the
(Fig. 2). This indicates that the partial treatment with LIPUS obliteration of the fracture line at the fracture gap in all
during Ph-1, Ph-2, or Ph-3 also is able to improve the groups (Figs. 3B–3E) was more advanced than that in the
mechanical properties of the fracture callus, as it does with control femur (Fig. 3A).
treatment throughout the 24 days. Furthermore, the maximal
The fracture line of the LIPUS-treated femur was bridged
torque for the LIPUS-treated side in the T group (193 Ϯ 39 by new bone formation 3D in all groups (Fig. 4). In contrast,
N-mm) was significantly higher than that for the other three bridging of the fracture site by new bone formation did not
groups: Ph-1, 170 Ϯ 39 N-mm; Ph-2, 159 Ϯ 31 N-mm; and occur substantially in the contralateral LIPUS-untreated
Ph-3, 161 Ϯ 34 N-mm (Fig. 2A). The data for stiffness was control femur. In the LIPUS-treated femur, the bridging of
almost the same as that for maximal torque (Fig. 2B). These newly formed cortex and trabecular bone in the fracture
results show that the continuous treatment with LIPUS was callus in the T group was greater than that in the Ph-1, Ph-2,
more effective toward the improvement of the mechanical and Ph-3 groups. These data show that the beneficial effect
properties of the fracture callus than the partial treatment.
of LIPUS treatment on the mechanical properties of fracture

ULTRASOUND ACCELERATES RAT FEMORAL FRACTURE REPAIR
675
in each group was independent of the size and bone mass of
the callus.
Histological evaluation
To examine the effects of LIPUS treatment on fracture
healing of the rat femur at the tissue and/or cellular level, we
used three rats per group (various durations and timings of
LIPUS treatment) for the histological evaluation. This his-
tological evaluation was a qualitative assessment. Histology
of LIPUS-treated femurs in the T group on days 9, 17, and
25, and in the Ph-2 and Ph-3 groups on day 17 and day 25
after the fracture are shown in Figs. 8–10.
Early endochondral ossification on day 9 in the LIPUS-
treated femur (Fig. 8B) was greater than that in the control
(Fig. 8A). In this case, mature cartilage tissue was observed
on both the control side and the LIPUS-treated side, but the
numbers of the hypertrophic chondrocytes and chondro-
clasts near the boundary between the cartilage tissue and
trabecular bone were greater in the LIPUS-treated side than
in the control side. These results suggest that LIPUS treat-
ment in the earlier period (Ph-1) give more advanced car-
tilage formation and early endochondral ossification in the
LIPUS-treated femur than in the control on day 9 after
fracture.
On day 17, endochondral ossification in the LIPUS-
treated femur (Figs. 9B and 9C) tended to be higher than
that in the control (Fig. 9A). The LIPUS-treated femur in
the Ph-2 group (Fig. 9B) on day 17 had received LIPUS
treatment from day 9 to 16 (for 8 days), and that in the T
group (Fig. 9C) on day 17 from day 1 to 16 (for 16 days)
after the fracture. Early remodeling of trabecular bone in the
hard callus also was observed in all femurs. However, the
size of the mature cartilage in the LIPUS-treated femur was
smaller than that in the control femur. The number of
osteoclasts on the trabecular bone surface in the hard callus
also was higher in the LIPUS-treated femur than that in the
control.
FIG. 3. Radiography of LIPUS-treated femur and nontreated femur
in each group at day 25 after the fracture. (A) Nontreated femur and
(B–E) LIPUS-treated femur, which are (B) Ph-1, (C) Ph-2, (D) Ph-3,
and (E) T, respectively. The femora were collected on day 25. All
femora including intramedullary pins were radiographed with a soft
X-ray system before dissection of soft tissue.
callus in each group is based on the improvement of the
fracture union.
Interestingly, more extensive bone bridging at the frac-
ture site was observed in the LIPUS-treated femur (Figs.
10B and 10C) than in the control femur (Fig. 10A) in not
only the T group but also in the Ph-3 group on day 25. Bone
HCA and BMC at fracture site
To determine the effects of LIPUS on the size of the bridging by new bone formation had not occurred substan-
fracture callus, we examined the outer HCA of the fractured tially in the LIPUS-untreated control femur (Fig. 10A). The
femur (Fig. 5A). Mean values of the HCA of femoral LIPUS-treated femur in the Ph-3 group (Fig. 10B) on day 25
fracture callus were approximately 20 mm2 in all groups on had received LIPUS treatment from day 17 to 24 (for 8
day 25 after the fracture (Fig. 6). There was no significant days), and that in the T group (Fig. 10C) on day 25, from
difference in the HCA between LIPUS-treated femurs and day 1 to 24 (for 24 days) after the fracture. In the LIPUS-
contralateral control femurs in any of the groups (Fig. 6).
treated femur, the bridging of newly formed cortex and
Because LIPUS did not affect the size of callus by 25 trabecular bone in the fracture callus was greater in the T
days after the fracture, we measured BMC at the fracture group than in the Ph-3 group.
site to determine the bone mass in the callus (Fig. 5B).
Values for BMC at the femoral fracture site on day 25 after
DISCUSSION
the fracture are shown in Fig. 7. Mean values of the BMC
in all groups were equivalent and approximately 0.028
g/area on day 25. There were no significant differences in
Fracture healing can be accelerated by exposure to LIPUS
BMC between the LIPUS-treated side and the contralateral in both animal fracture models and clinical trials.(3–9) How-
control side of the femoral fracture site in any of the groups. ever, the underlying mechanism of action remains unclear.
These data suggest that the beneficial effect of LIPUS In this study, we investigated the effects of the duration and
treatment on the mechanical properties of the fracture callus timing of LIPUS treatment on the benefits of LIPUS on the

676
AZUMA ET AL.
FIG. 4. 3D reconstructions from 100 ␮CT to-
mograms (each 80 ␮m in thickness) of fractured
femora treated with LIPUS or nontreated on day
25 after the fracture. Reconstructions of non-
treated femur in the T group and LIPUS-treated
femurs in Ph-1, Ph-2, Ph-3, and T groups are
shown.
healing of rat femoral fractures to find a target reaction of showed that the density and connectivity of bone in the
LIPUS in the fracture healing process. fracture callus was important to increase the rigidity of the
Maximal torque of the LIPUS-treated side was signifi- callus,(24–26) our observations in this study suggest that the
cantly higher than that of the contralateral control side in all efficiency of fracture union also contributes to the advanced
groups. These results show that partial LIPUS treatment is mechanical properties of the fracture callus. Here, we de-
able to accelerate fracture healing regardless of its timing. scribe for the first time that the multiple viewing from
Furthermore, maximal torque of the LIPUS-treated femur in various angles (60° pitch) using 3D ␮CT reconstructions of
the T group was significantly higher than that in the other the fracture callus is a powerful tool for estimation of the
three groups, probably because the accelerating effect of bone union and the fracture callus condition both visually
LIPUS was additive. These data also suggest that LIPUS and quantitatively.
acts on some kind of cellular reaction involved in the
fracture healing process.
In the histological analysis, endochondral ossification on
day 9 and day 17, and the bone remodeling and bone
The mechanical properties (i.e., torsional torque and stiff- bridging on day 17 and day 25 had advanced more in
ness) of the fracture callus in long bones reflect the extent of LIPUS-treated femurs than in the control femurs, respec-
fracture union as a result of healing in the fracture site.(22,23) tively. These observations assured the results from the me-
In this study, maximal torque of the LIPUS-treated femur chanical testing and multiple viewing by ␮CT. The partial
was significantly greater than that of the control side in all LIPUS treatments (8-day treatment) only in Ph-1, Ph-2, or
groups, but there was no significant difference between the Ph-3 were effective in advancing endochondral ossification
LIPUS-treated side and the control side for BMC at the and/or bone remodeling in the fracture callus. These results
femoral fracture site or for the HCA in any of the groups. suggest that LIPUS influenced some cellular reactions ac-
These findings indicate that the promoting effects of LIPUS tive at each time period. However, it is unclear which
on the mechanical properties of the fracture callus involve reaction is the target of LIPUS and how LIPUS affects it.
the fracture union but not the size and/or bone mass of the
fracture callus on day 25 after the fracture.
The micromechanical strains produced by LIPUS’s pres-
sure waves in living tissue can result in biochemical events
The results for the multiple viewing of 3D ␮CT recon- at the cellular level. In accordance with this, it seems to be
structions at the fracture site were coincident with the data easy that a cell type that has a sensitivity to mechanical
for mechanical testing. Although several investigators strains would respond to LIPUS. Although further investi-

ULTRASOUND ACCELERATES RAT FEMORAL FRACTURE REPAIR
677
FIG. 7. Effects of LIPUS treatment of various durations and timings
on BMC of the fracture callus near the fracture site on day 25 after the
fracture. The HCA of the nontreated femur and LIPUS-treated femur is
represented as open bars and closed bars, respectively. There was no
significant difference in BMC between the nontreated femur and the
LIPUS-treated femur in the T group (n ϭ 11), Ph-1 group (n ϭ 9), Ph-2
group (n ϭ 12), or Ph-3 group (n ϭ 12). The horizontal bars repre-
sent Ϯ 1SD. Paired t-testing was used for statistical analysis.
FIG. 5. (A) The BMC of femoral diaphysis at fracture site was
measured, and (B) anterior and posterior HCAs were traced and mea-
sured with image analysis software.
FIG. 8. Effects of LIPUS treatment on the histology of fracture
healing at day 9. Photomicrographs show the appearance of the fracture
healing in (A) the control limb and (B) the treated limb. Early endo-
chondral ossification on day 9 in the LIPUS-treated femur was greater
than in the control. Sections were stained with Masson’s trichrome
stain. Arrowheads show the fracture sites. Specific regions are labeled
as follows: cortical bone, co; muscle, m; intramedullary canal, im; and
cartilage, c.
FIG. 6. Effects of LIPUS treatment of various durations and timings
on the HCA of the fracture callus on day 25 after the femoral fracture.
The HCA of nontreated femur or LIPUS-treated femur is represented as
open bars or closed bars, respectively. There was no significant differ-
ence in the HCA between the nontreated femur and LIPUS-treated
femur in the T group (n ϭ 11), Ph-1 group (n ϭ 9), Ph-2 group (n ϭ
12), or Ph-3 group (n ϭ 12). The horizontal bars represent Ϯ1 SD.
Paired t-testing was used for statistical analysis.
potassium channel, the activity of which was modified
by mechanical stress.(27,28) Furthermore, it was reported
that high-energy ultrasound modulated the activity of
thymocytes.(29)
gations are needed to define the target cells in each time
period, the results obtained in the present study allow us to
make the following speculation about the candidate:
(2) Blood flow and neovascularization. After or overlap-
ping the end stage of inflammatory reactions, the for-
mation of granulation tissue and the invasion of multi-
potential mesenchymal stem cells occur. In this stage,
the concurrent proliferation of blood vessels within the
periosteal tissues and marrow space helps to route the
appropriate cells to the fracture site. Rawool et al.
(1) Inflammatory reaction. In the inflammatory phase, a
hematoma is formed from blood vessels ruptured by the
injury; and soon thereafter, inflammatory cells such as
macrophages and/or lymphocytes invade the clot. It was
reported that macrophages contain a stretch-sensitive

678
AZUMA ET AL.
FIG. 9. This shows the effect of 8 days of treatment versus 16 days
of treatment. On day 17, endochondral ossification and early remodel-
ing were less in the (A) control femur than in the (B) 16 days of
LIPUS-treated femur or (C) 8 days of LIPUS-treated femur. The size of
the mature cartilage in the LIPUS-treated femur was smaller and the
number of osteoclasts was more in the LIPUS-treated femora than in
the control.
FIG. 10. On day 25, (A) the control femur had less extensive bone
bridging than the femur treated for (B) 24 days or (C) 8 days. The
bridging of newly formed cortex and trabecular bone in the fracture
callus was greater in panel B than in panel C.
intramembranous ossification.(32) Ryaby et al. reported
that LIPUS increased calcium incorporation in differenti-
ating cartilage and bone cell cultures.(33,34) This increased
second messenger activity was paralleled by the modula-
tion of adenylate cyclase activity and transforming growth
factor ␤ (TGF-␤) synthesis in osteoblastic cells.(34) Fur-
thermore, other investigators reported that topical applica-
tion of TGF-␤ to long bones stimulated osteogenesis in the
periosteum.(35–37)
showed that LIPUS increased the degree of vascularity
in an osteomized dog ulna model of fracture healing.(30)
This effect was observed from 1 to 3 days after the
fracture to over 10 days thereafter, and we showed that
LIPUS stimulated platelet-derived growth factor
(PDGF) production in the coculture human endothelial
cells with human osteoblasts in vitro.(31)
(4) Chondrogenesis. Parvizi et al. reported that ultrasound
modulated second messenger activity in primary rat
chondrocytes.(38) They showed that application of ul-
(3) Intramembranous ossification. Periosteum-derived cells,
which are of the osteoblastic lineage, may be involved in

ULTRASOUND ACCELERATES RAT FEMORAL FRACTURE REPAIR
679
trasound at 50 mW/cm2 and at higher intensities in-
duced a real-time increase in the intracellular level of
calcium. And other investigators showed that exposure
of cultured chondrocytes to LIPUS stimulated an up-
regulation of aggrecan gene expression.(39,40)
7. Wang S-J, Lewallen DG, Bolander ME, Chao EYS, Ilstrup
DM, Greenleaf JF 1994 Low intensity ultrasound treatment
increases strength in a rat femoral fracture model. J Orthop Res
12:40–47.
8. Heckman JD, Ryaby JP, McCabe J, Frey JJ, Kilcoyne RF 1994
Acceleration of tibial fracture-healing by non-invasive, low-
intensity pulsed ultrasound. J Bone Joint Surg Am 76:26–34.
9. Kristiansen TK, Ryaby JP, McCabe J, Frey JJ, Roe LR 1997
Accelerated healing of distal radial fractures with the use of
specific, low-intensity ultrasound. A multicenter, prospective,
randomized, double-blind, placebo-controlled study. J Bone
Joint Surg Am 79:961–973.
10. Binderman I, Zor U, Kaye AM, Shimshoni Z, Harell A,
Somjen D 1988 The transduction of mechanical force into
biochemical events in bone cells may involve activation of
phospholipase A2. Calcif Tissue Int 42:261–266.
11. Buckley MJ, Banes AJ, Levin LG, Sumpio BE, Sato M, Jordan
R, Gilbert J, Link GW, Tran Son Tay R 1988 Osteoblasts
increase their rate of division and align in response to cyclic,
mechanical tension in vitro. Bone Miner 4:225–236.
12. Rubin J, Biskobing D, Fan X, Rubin C, McLeod K, Taylor WR
1997 Pressure regulates osteoclast formation and MCSF ex-
pression in marrow culture. J Cell Physiol 170:81–87.
13. Bolander ME 1992 Regulation of fracture repair by growth
factors. Proc Soc Exp Biol Med 200:165–170.
(5) Endochondral ossification. Invasion of newly formed
blood vessels into the cartilage tissue, degradation of
cartilage tissue, and bone formation are involved in
endochondral ossification. The growth, differentiation
and activity of the cells such as chondrocytes, osteo-
blasts, endothelial cells, and chondroclasts/osteoclasts
related to those reactions may be important in this
process. In a previous study, we showed that the num-
ber of multinuclear cells (MNCs) near the boundary
between cartilage tissue and trabecular bone in the
LIPUS-treated femur (15 Ϯ 3 MNCs/mm) was greater
in comparison with that in the control femur (8 Ϯ 2
MNCs/mm) on day 21 after the fracture (in prepara-
tion). However, it is unclear whether this effect was
based on the direct action of LIPUS on osteoclastogen-
esis. Soma et al. showed that the application of bilateral
stretch stress to osteoblastic cells stimulated the produc-
tion of some factors that could up-regulate osteoclasto- 14. Schwachtgen JL, Houston P, Campbell C, Sukhatme V, Brad-
genesis.(41)
dock M 1998 Fluid shear stress activation of egr-1 transcrip-
tion in cultured human endothelial and epithelial cells is me-
(6) Bone remodeling. Tanzer et al. reported that LIPUS
diated via the extracellular signal-related kinase 1/2 mitogen-
stimulated bone ingrowth into porous coated dog fem-
activated protein kinase pathway. J Clin Invest 101:2540–
oral implants.(42) In contrast, Spadaro et al. reported that
2549.
physeal bone growth was far less sensitive to LIPUS
15. MacKenna DA, Dolfi F, Vuori K, Ruoslahti E 1998 Extracel-
than was fracture repair.(43) It remains unclear whether
lular signal-regulated kinase and c-Jun NH2-terminal kinase
and/or how LIPUS directly modulates bone metabolism
involved in bone formation and bone resorption.
activation by mechanical stretch is integrin-dependent and
matrix-specific in rat cardiac fibroblasts. J Clin Invest 101:
301–310.
In conclusion, we investigated whether the accelerating
effects of LIPUS on the healing of rat femoral fractures
depended on the duration and timing of LIPUS treatment to
determine the target reaction of LIPUS in the fracture heal-
ing process. Although we could not define the main target,
we found out that LIPUS accelerated the rat femoral frac-
ture healing regardless of the timing of LIPUS treatment.
Thus, LIPUS appears to act on various cellular reactions
involved in the fracture healing process.
16. Pavalko FM, Chen NX, Turner CH, Burr DB, Atkinson S,
Hsieh YF, Qiu J, Duncan RL 1998 Fluid shear-induced me-
chanical signaling in MC3T3–E1 osteoblasts requires
cytoskeleton-integrin interactions. Am J Physiol 275:C1591–
C1601.
17. Quinn TM, Grodzinsky AJ, Buschmann MD, Kim YJ, Hun-
ziker EB 1998 Mechanical compression alters proteoglycan
deposition and matrix deformation around individual cells in
cartilage explants. J Cell Sci 111:573–583.
18. Ajubi NE, Klein-Nulend J, Alblas MJ, Burger EH, Nijweide
PJ 1999 Signal transduction pathways involved in fluid flow-
induced PGE2 production by cultured osteocytes. Am J
Physiol 276:E171–E178.
REFERENCES
19. Einhorn TA 1998 The cell and molecular biology of fracture
healing. Clin Orthop 55(Suppl 3):S7–S21.
20. Jingushi S, Joyce ME, Bolander ME 1992 Genetic expression
of extracellular matrix proteins correlates with histologic
changes during fracture repair. J Bone Miner Res 7:1045–
1055.
21. Bonnarens F, Einhorn TA 1984 Production of a standard
closed fracture in laboratory animal bone. J Orthop Res 2:97–
101.
22. White AA III, Panjabi MM, Southwick WO 1977 The four
biomechanical stages of fracture repair. J Bone Joint Surg Am
59:188–192.
1. Ziskin MC 1987 Application of ultrasound in medicine—
comparison with other modalities. In: Rapacholi MH, Gran-
dolfo M, Rindi A (eds.) Ultrasound: Medical Applications,
Biological Effects, and Hazard Potential. Plenum Press, New
York, NY, USA, pp. 49–59.
2. Wells PNT 1985 Surgical applications of ultrasound. In: Ny-
borg WL, Ziskin MC (eds.) Biological Effects of Ultrasound.
Churchill Livingstone, New York, NY, USA, pp. 157–167.
3. Xavier CAM, Duarte LR 1983 Estimulaci ultra-sonica de callo
osseo: Applicaca clinica. Rev Bras Orthop 18:73–80.
4. Duarte LR 1983 The stimulation of bone growth by ultra-
sound. Arch Orthop Trauma Surg 101:153–159.
23. Black J, Perdigon P, Brown N, Pollack SR 1984 Stiffness and
strength of fracture callus. Relative rates of mechanical mat-
uration as evaluated by a uniaxial tensile test. Clin Orthop
182:278–288.
24. Markel MD, Wikenheiser MA, Chao EY 1990 A study of
fracture callus material properties: Relationship to the tor-
sional strength of bone. J Orthop Res 8:843–850.
5. Klug W, Franke WG, Knoch HG 1986 Scintigraphic control of
bone-fracture healing under ultrasonic stimulation: An animal
experimental study. Eur J Nucl Med 11:494–497.
6. Pilla AA, Mont MA, Nasser PR, Khan SA, Figueiredo M,
Kaufman JJ, Siffert RS 1990 Non-invasive low-intensity
pulsed ultrasound accelerates bone healing in the rabbit. J Or-
thop Trauma 4:246–253.

680
AZUMA ET AL.
25. Chakkalakal DA, Lippiello L, Wilson RF, Shindell R, Con- 37. Taniguchi Y, Tanaka T, Gotoh K, Satoh R, Inazu M 1993
nolly JF 1990 Mineral and matrix contributions to rigidity in
fracture healing. J Biomech 23:425–434.
26. Augat P, Merk J, Genant HK, Claes L 1997 Quantitative
assessment of experimental fracture repair by peripheral com-
puted tomography. Calcif Tissue Int 60:194–199.
27. Martin DK, Bootcov MR, Campbell TJ, French PW, Breit SN
1995 Human macrophages contain a stretch-sensitive potas-
sium channel that is activated by adherence and cytokines. J
Membr Biol 147:305–315.
28. Mattana J, Sankaran RT, Singhal PC 1995 Repetitive mechan-
ical strain suppresses macrophage uptake of immunoglobulin
G complexes and enhances cyclic adenosine monophosphate
synthesis. Am J Pathol 147:529–540.
29. Dooley DA, Child SZ, Carstensen EL, Miller MW 1983 The
effects of continuous wave and pulsed ultrasound on rat thy-
mocytes in vitro. Ultrasound Med Biol 9:379–384.
30. Rawool NM, Goldberg BB, Forsberg F, Winder AA, Tallah
RJ, Hume E 1997 Power doppler assessment of vascular
changes during fracture treatment with low intensity ultra-
sound. Trans 83rd Radiol Soc North Am 83:421.
31. Ito M, Azuma Y, Ohta T, Komoriya K 2000 Effects of ultra-
sound and 1,25-dihydroxyvitamin D3 on growth factor secre-
tion in co-cultures of osteoblasts and endothelial cells. Ultra-
sound Med Biol 26:161–166.
Transforming growth factor beta-1 induced cellular heteroge-
neity in the periosteum of rat parietal bones. Calcif Tissue Int
53:122–126.
38. Parvizi J, Parpura V, Kinnick RR, Greenleaf JF, Bolander ME
1997 Low intensity ultrasound increases intracellular concen-
tration of calcium in chondrocytes. Trans Orthop Res Soc
22:465.
39. Yang K-H, Parvizi J, Wang S-J, Lewallen DG, Kinnick RR,
Greenleaf JF, Bolander ME 1996 Exposure to low-intensity
ultrasound increases aggrecan gene expression in a rat femur
fracture model. J Orthop Res 14:802–809.
40. Wu C-C, Lewallen DG, Bolander ME, Bronk J, Kinnick R,
Greenleaf JF 1996 Exposure to low intensity ultrasound stim-
ulates aggrecan gene expression by cultured chondrocytes.
Trans Orthop Res Soc 21:622.
41. Soma S, Matsumoto S, Takano-Yamamoto T 1997 Enhance-
ment by conditioned medium of stretched calvarial bone cells
of the osteoclast-like cell formation induced by parathyroid
hormone in mouse bone marrow cultures. Arch Oral Biol
42:205–211.
42. Tanzer M, Harvey E, Kay A, Morton P, Bobyn JD 1996 Effect
of noninvasive low intensity ultrasound on bone growth into
porous-coated implants. J Orthop Res 14:901–906.
43. Spadaro JA, Albanese SA 1998 Application of low-intensity
ultrasound to growing bone in rats. Ultrasound Med Biol
24:567–573.
32. Utvag SE, Grundnes O, Reikeras O 1999 Early muscle-
periosteal lesion inhibits fracture healing in rats. Acta Orthop
Scand 70:62–66.
33. Ryaby JT, Bachner EJ, Bendo JA, Dalton PF, Tannenbaum S,
Pilla AA 1989 Low intensity pulsed ultrasound increases cal-
cium incorporation in both differentiating cartilage and bone
cell cultures. Trans Orthop Res Soc 14:15.
34. Ryaby JT, Mathew J, Duarte-Alves P 1992 Low intensity
pulsed ultrasound affects adenylate cyclase and TGF-␤ syn-
thesis in osteoblastic cells. Trans Orthop Res Soc 17:590.
35. Joyce ME, Roberts AB, Sporn MB, Bolander ME 1990 Trans-
forming growth factor-beta and the initiation of chondrogen-
esis and osteogenesis in the rat femur. J Cell Biol 110:2195–
2207.
Address reprint requests to:
Yoshiaki Azuma, Ph.D.
Pharmacological Research Department
Teijin Institute for Biomedical Research, Teijin Limited
4–3-2 Asahigaoka, Hino City
Tokyo 191-8512, Japan
36. Joyce ME, Jingushi S, and Bolander ME 1990 Transforming
growth factor-beta in the regulation of fracture repair. Orthop
Clin North Am 21:199–209.
Received in original form July 30, 1999; in revised form October
31, 2000; accepted October 31, 2000.