o

Q.
(D
OH

u
D
Ultrasound Produced by a
CD
to Conventional Therapeutic Ultrasound
0
Unit Accelerates Fracture Repair
Background and Purpose. A recent novel appliciition of ultrasotind therapy is
the treatment of bone fractures. The aim of this study was to investigate the
effect on fracture repair of ultrasound produced by a conventional tlieraptni-
tic ultrasound luiit as used by physical therapists. Subjects and Methods.
Bilateral midshaft femur fractures were created in 30 adult male Long-Evans
rats. Ultrasound therapy was commenced on the first day after fracture and
introduced 5 days a week (or 20 minutes a day. Each animal was treated
unilaterally with active ultrasound and contralaterally with inactive ultra-
sound. Active ultrasound involved a 2-millisecond hurst of 1.0-MHz sine waves
repeating at 100 Hz. The spatially averaged, temporally averaged intensity was
set at 0.1 W/cm^. Animals were killed at 25 and 40 days after fracture
induction, and the fractures were assessed for hone mass and strength. Results.
There were no differences between fractures treated with active ultrasound
and fractures treated with inactive ultrasound at 25 days. However, at 40 days,
active ultrasound-treated fractures had 16.9% greater bone mineral content at
the fracture site than inactive ultrasound-treated fractures. This change
resulted in a 25.8% increase in bone size, as opposed to an increase in bone
density, and contributed to active ultrasound-treated fractures having 81.3%
greater mechanical strength than inactive ultrasound-treated fractures, Dis-
cussion and Conclusion. These data indicate that ultrasound produced by a
conventional therapeutic ultrasound unit as traditionally used by physical
therapists may be used to facilitate fracture repair. However, careful interpre-
tation of this controlled laboratoiT study is warranted until its findings are
confirmed by clinical trials. [Warden SJ, Fuchs RK, Kessler CK, etal. Ultra-
sound produced hy a conventional therapeutic ultrasound unit accelerates
fracture repair. Phys Ther. 2006;86:l 118-1127.]

K e y W o r d s : Animal, Bone, Fractures, Models, Musculosheletal diseases, Orth(>f)edic equipment, Oiihoj)edic

procedures, Physical therapy techniques. Skeleton, Sports medicine.

Stuart J Warden, Robyn K Fuchs, Chris K Kessler, Keith G Avin, Ryan E Cardinal, Rena L Stewart

1118 Physical Therapy . Volume 86 , Number 8 , August 2006

 ltrasound is a form of acoustic energy (sound) interfacial effects at the bone surface resulting from

U that possesses a frequency above the limit

detectahle hy the hiunan ear. It has been used
therapeutically for more than half a centuiy
and currently is one of the most widely and frequently
used electro therapeutic modalities applied by physical
therapists.' However, its full therapeutic potential is far
from established, with new applications being added
regularly to its repertoire.'' One recent novel application
is in the treatment of bone fractures.-'*'*
Fracture repair involves a complex cascade of cellular
events incorporating appropriate cellular recruitment,
timed genetic expression, and the sequenced synthesis
of numerous compounds. Although it is considered to
be a naturally optimized process, recent evidence has
demonstrated that fracture repair can he infiuenced by
ultrasound to occtxr more rapidly without compro-
mising the final tissue-level outcome."^-^ The application
of ultrasound in animal fracture models accelerated
the return to mechanical strength of intact bone by 30%
to ;18%.'^ Similarly, in well-designed clinical trials,
ultrasound accelerated the rate of fracture repair in
the tibia, radius, and scaphoid by 30% to 38%.*^-^ By
pooling of the clinical trial data, a weighted average
effect size was calculated to be 6.41 (95% confidence
inten^al [CI] =0.01-11.81); this value converts into a mean
improvement in healing time of 64 days with ultrasound.^
The results of studies of the effect of ultrasound on
fractured bone are interesting from the perspective
that physical therapists tradidonally have been in-
structed to avoid the application of ultrasonic energy to
bone. When ultrasound is applied to bone, there is an
inherent risk of tissue damage. Ultrasound has selective
bone ha\ing a high absorption coefficient, a high rela-
tive acoustic impedance, and an ability to propagate
shear waves.'" When doses at the high end of the
therapeutic range are used, these effects can generate
considerable tissue damage attributable to heating and
inertial cavitadon effects."-^^ To achieve clinically sig-
nificant improvements during fracture repair, without
the risk for tissue damage, the ultrasound dose has been
changed substantially from that traditionally introduced
in physical therapist clinical practice. Clinically, ultra-
sound is introduced at an intensity commonly in the
range of 0.5 to 2.0 W/cm'"^.' In comparison, in investi-
gations into the therapeutic effect of ultrasound on
bone, low-intensity pulsed ultrasoimd (LIPUS) has been
used. Low-intensity pulsed ultrasound is pulsed-wave
ultrasound with a spatially averaged, temporally aver-
aged intensity of iess than 0.1 W/cm^.'' With LIPUS,
heat generation at the soft tissue-bone interface has
been shown both theoretically'^ and experimentally'' to
he insignificant (<1.0°C). Similarly, the risk for tissue-
damaging inertial cavitation is negligible.'*
Although LIPUS has been found to be effective in the
management of bone fractures, to date the clinical utility
of this finding in physical therapy is limited. Special-
ized ultrasound units (Exogen 2000+*) have been de-
veloped for the treatment of fractured hone. Although
these units are highly efficacious,"'-'^ their cost is prohib-
itive because the units are leased on a patient-to-patient
basis rather than purchased by individual clinics. Despite
the benefits observed with LIPUS, the high cost of the
* Smith & Ncphfw, Orthopaedic Division, 14.W Brooks Rd, Memphis. TN 381 !fi.

SJ Warden, PT, PhD, is Assistant Professor, Department of Physical Therapy and Department of Anatomy and Cell Biology, Indiana University, 1140
W Michigan St, CF-326, Indianapolis, IN 46202 (USA). Addres.s all correspondence to Dr Warden at: stwarden@iupui,edti.

RK Fuchs. PhD, is A.ssistant Research Professor, Department of Anatomy and Cell Biology, Indiana University.

CK Kessler, BS, is Research Assistant. Department of Physical Therapy, Indiana University, He was completing his MD studies at the Indiana
University School of Medicine at the lime of this study.

KG Avin, PT. DPT, is Research Assistant, Department of Physical Therapy, Indiana University. He was completing his DPT studies at ihe time of
this study.

RE Cardinal, PT, DPT, is Research A.s.sistant, Department of Physical Therapy, Indiana University. He was completing his DPT studies al the lime
of this study.

RL Stewart, MD, FRCS(C), is Director of Orthopaedic Trauma, Wishard Health Services, and Assistant Professor of Orthopaedic Surgery, Indiana
University School of Medicine, Indianapolis, Ind,

All authors provided concept/idea/research design, data collection, and con.sultation (including review of manuscript hefore submission). Dr
Warden, Dr Fuchs, and Dr Stewart provided writing. Dr Warden provided data analysis, project management, and fund procurement.

All procedures were performed with prior approval of the Institutional Animal Care and Use Committee of Indiana University.

Thk article was received November 16. 2005, and was accefjted Ff-bruary 15, 2006.

Physical Therapy . Volume 86 , Number 8 . August 2006 Warden et al , 1 119

 specialized ultrasotind units has led some authors'" to
question whether conventional therapeutic ultrasound
units could be used by physical therapists to accelerate
fracture repair. At the lower-intensity settings on these
units, it is possible to produce a dose comparable to tbat
sbown to be effective dining fracture repair witb the
specialized units.
The aim of this study was to investigate the effect of
LIPUS produced by a conventional therapeutic ultra-
sound unit on fracture repair in an animal model. We
bypothesi/ed tbat LIHUS would facilitate fiacture repair,
as evidenced by more bone mineral at tbe fracture site
and a stronger fracture calltis at selected time points
during healing.
Method
Animals
Thirty adult male Long-Evans rats {weight=350-400 g)
were purchased^ and acclimated for 1 week before
experimentation. Animals had ad libitum access to stan-
dard rat cbow and water at all times.
Fracture Induction
All animals underwent snrger\' upon entry into the studv
to create bilateral midshaft femur fractures. The fur was
clipped and cleaned with alternating chlorhexidine
and 70% ethanol scrubs. After a preoperative stibcuta-
neous dose of btiprenorpbine bydrochloride analgesia^
(0.05 mg/kg), surgical anesthesia was achieved with a
mixture of ketamine^ {60-80 mg/kg) and xylazine^
{7.5 mg/kg) introduced intraperitoneally. With a sterile
technique, a 25-mm longitudinal incision was made over
the lateral thigh, beginningjtist distal to the lateral knee
joint and extending proximally. The intermuscular sep-
tum between the vastus lateralis and the hamstring
muscles was divided by blunt dissection to localize the
femur. The lateral structures stabilizing the patella were
divided, and the patella was manually dislocated medi-
ally. Tbe femur was fractured at its midsbaft by means of
a transverse osteotomy with a Dreniel drill" having an
attached diamond-embedded wafer blade (Super Flex
Diamond Disc*). To stabilize the fracture, a l.(>mm-
diameter stainless steel K-wire** was in.serted retrograde
into the intiamedullaiy canal, beginning in the knee
between tbe femoral condyles and exiting the greater
trochanter. The pin was cut as clcse as possible to tbe
knee artictilar cartilage and driven proximally so that the
' Harlan .Spiasin--r)awlcy lm\ Pt) Biix 2917fi, [iidiatmpulis, IN 46229,
' Rcckitt Belli ki,'ic-r PhMrTiiaccmicals LiiJ, Inc, 10710 Midlotliian Tiinipikf, Stiiu
430, Ritliiiniiid. VA ^-lasr..
^ Foit Dwigc Aiiinial Health, HOII .5ih Si NW. Fun Dodge, L\ .50301,
" R()bfri Basrh Tool Corp, 1800 W Central Rd. Moinii Prosprtl. II. 00065.
' Millt-x Inc. 589 Davit-s Dr, York, PA 17402,
** Svdrlifs Inr, 1302 WriKlits l.n K. Wt-st Clifstci, PA I93HII.
1 20 . Warden et al
tip was flush with the cartilage. The patella was relocated
and stabilized with an absorbable suture, and absorbable
sutures were tised to close the intermuscular septum and
skin incision. Tbe procedure was repeated on the con-
tralateral side to create bilateral fractures.
Ultrasound Intervention
Ultrasound therapy commenced on the first day after
fracture induction. This starting time point is consistent
with previous studies'^ '•'-'" and tbe belief tbat ultrasound
influences early cellular processes immediately after
bone injury.'-'^'" Each animal was treated unilaterally
with active LIPUS and contialaterally witb inactive
LIPUS (placebo). For intervention, animals were anes-
thetized with inhalation of 3% isoflurane^^ at 1.5 L/min
in a plastic container and then with 1.5% isofluorane at
1.5 L/min via a face mask (for maintenance of anesthe-
sia). Active LIPUS was produced with an Accusonic
LIPUS GS 170 ultrasound unit/^ which produces a
2-miilisecond burst of 1.0-MHz sine waves repeating at
100 Hz. The spatially averaged, temporally averaged
intensity on tbis unit is set at 0.1 W/cm'"', which repre-
sents the average ultrasound output over the area of tbe
ultrasound beam {spatial average) and tbe average of
tbis intensity over a cotnplcte pulse cycle (ultrasound
"on" and "off" periods; temporal average). Tbe manu-
facturer reported that tbe transducer bad an effective
radiating area of 5 cm^ and a beam nonuniformity ratio
of less than 6.0. Ultrasound unit performance was con-
firmed at weekly intervals witb a power meter (UPM-DT-
1^^). Active LIPUS and inactive LIPUS were coupled
with the skin by use of ultrasound gel (Aquasonic lOO""
and introduced 5 d/wk for 20 min/d by use of a
stationaiy treatment head. The fur was clipped at weekly
intenals to facilitate ultrasound propagation. The
LIPUS parameters and treatment time were cliosen on
the basis of those shown to be beneficial during the
healing of tissue injuries (reviewed by Warden''').
Assessment Time Points
All animals were evaluated intraoperatively and 1 week
postoperatively to assess tbe rotatoiy stability of the
fractures. Animals with a fracture tbat was rotationally
tmstable at postoperative week I, indicating inadequate
fracture site fixation, were excluded from tbe study. All
other animals were killed at 25 and 40 days after fracture
induction by inhalation of anesthetic gases followed by
cemcal dislocation. Animals in tbe 25- and 40-day
groups received 16 and 27 LIPUS treatments, respec-
tively. After death, the left and right femurs were har-
^^ Abbott I .iihoraiorics. 100 Abbott Park Rtl, Abbott Park. TL fiOO64.
^^ Mflron Medical .Xiwralia Ply Ltd. PO Bo\ 21 tifi, Carnini Down.s, Victoria .^2
Australia,
^^ Ohmic Insirutitcms, ".08 Aiigtist St, E^Lstort, MD 2111(11.
"" Parker l.aboratoricN Iiif, 28fi Eldridge Rd, Fairfield. N| 07004.
Physical Therapy , Volume 86 , Number 8 . August 2006
vested, wrapped in saline-soaked gauze, and stored at
Radiography
Postmortem ex vivo radiographs were obtained with a
specimen radiography system.** The femurs were posi-
tioned for both anteroposterior and lateral radiographs
on dental film (Kodak Ultraspeed Dental Film [size
4]***). Samples were exposed to a voltage of 18 kV for
10 seconds. After film processing, the stage of fracture
healing was qtiantified with a 4-point radiographic scor-
ing system (O^no evidence of healing; 1-callus forma-
tion evident but fracture gap not yet bridged; 2=callus
formation evident with possible bridging of the fracture
gap; and 3~fracture union). The examiner was unaware
of both femur side and time since fracture during
grading.
Microcomputed Tomography
The stabilizing steel K-wires were carefully removed from
the intramedullaiy canal before further assessment, as
metal causes beam-hardening artifacts during quantita-
tive radiograpbic imaging. Microcomputed tomography
was performed on a randomly selected subgroup of
fractures to visualize 3-dimensionally the stage of frac-
ture healing at 25 and 40 days. Each femur was placed in
a 13.S-mm-diameter plastic tube filled with saline and
centered in the gantiy of a desktop microcompnted
tomography machine (p.CT-20^^^). A scout scan was
perfonned to enable fracture site localization, and 2B0
slices were taken with an isotropic voxel size of 26 (xm
and an integration time of 150 milliseconds. A standard
convohuion-back projection procedure with a Shepp-
Logan filter was used to reconstruct the computed
tomography images in 1.024 X 1,024-pixel matrices.
Dual-Energy X-ray Absorptiometry
Dual-energy X-ray absorptiometry (DXA) was performed
to assess fracture site bone mineral content (BMC, in
grams). Femurs were positioned on their caudal surface
on a mouse densitometer (PlXlmus^^*) with iiltrahigh
resolution (0.18X0.18 mm per pixel). Left and rigbt
femur pairs from each animal were scanned side by side
on the same scan. Upon completion of each scan, a
mutually exclusive region of interest (18X10 mm) was
centered over each fracture site.
Peripheral Quantitative Computed Tomography
Peripheral quantiuuive computed tomography (pQCT)
was used to assess fracture site volumetric bone mineral
density (vBMD, in milligrams per cubic centimeter),
"" Faxitroii X-ray Corp, 22!i l.arkln Dr. Unit 1, Whrrlicig. IL WOW.
*** Eastman Kodak Co. 343 Slate St. Rochfslcr. NY liGM.
" ' Scan(() Medical ACl, Auenrins fi-H, ^^303 Bassersdorf. .Switzerland.
=" Lunar Corp. 313 W Beltline Hgwy. Madison, WI 537t4.
Physical Therapy , Volume 86 , Number 8 , August 2 0 0 6
BMC (in milligrams per millimeter), and bone area
(B.Ar, in sqtiare millimeters). Each femur was placed in
a plastic tube Jilled with saline and centred in the gantr)'
of a pQCT machine (XCT Research SA+^^^). After a
scout view was obtained to enable scan localization, 5
cross-sectional scans were obtained with a H)-fxm voxel
size. The middle scan was centered through the frac-
ture line, and the other scans were positioned 1.5 mm
and 3.0 mm above and below the center scan. During
analyses, the bone edge was detected with contour mode
1 at a threshold of 400 mg/cm"' within the Stratec
software.^^^ Tbe data for the 5 slices per bone were
averaged.
Destructive Mechanical Testing
Tbe mechanical properties of the fracture site were
assessed by testing the femurs in 4-point bending
(Fig. lA). Bones were slowly brought to room tempera-
ture overnight in a saline bath. Femurs were positioned
cranial side up across the lower contacts of a custom-
built 4-point bending rig on an Alliance RT/5 Materials
Testing System.""" The lower contacts had a span width
of 20.0 mm. The upper contacts were pivoted to en-
stire that both contacts simultaneously touched tbe
cranial surface of the bone when the cross head was
lowered. The upper contacts had a span width of
8.0 mm, centered between tbe lower contacts. Tbe
upper contacts were lowered to fix the bone in place
with a 1.0 N static preload. Tbe bone was subsequently
broken in 4-point bending with a cross-head speed ot
20.0 mm/min. During testing, force and displacement
data were collected every 0.1 second (at a frequency of
10 Hz) with TestWorks 4 software (version 4.08A)."""
Force-displacement curves were visually inspected, and
ultimate force (in newtons), stiffness (in newtons per
milhmeter), and energy- to ultimate force (in millijoules)
were recorded (Fig. IB).
Data Analysis
Statistical analyses were performed with the Statistical
Package for Social Sciences software,"** with a level of
significance set at .05 for all tests. The significance of
radiograpbic scores was determined with the Wilcoxon
signed rank test, whereas fracture site bone mass and
mechanical properties were compared with paired t
tests. Ultrasound intei-vention (active LIPUS versus inac-
tive LIPUS) was the within-animal independent variable
for all tests. In addition, effect sizes on fracture site bone
mass and mechanical properties were determined with
mean percent differences (%diff) and 95% Cls of the
*^^ Slratec Mfdi/iiuechiiik (Inihll. Diirlaclicr Siiassc 35. [)-7,'il72 Piorzhcim.
Germany,
"II" MTS Sysitrins Corp. 1400(1 Technology tJr, Eden Praifie. MN ^^
*** SPSS Inc. 2.3.S S V\'ackcr Dr. Chicago, It. 60606.
Warden etal . 1 1 2
 mean percent differences between active LIPUS-treated
fractures and inactive LIPUS-treated fractures.

Results

Animal Characteristics
One animal from the 40-day group died from surgical
complications during fracture induction. Three other
animals (1 animal and 2 animals from the 25-day and
40-day groups, respectively) were excluded at postoper-
ative week 1 because of rotatory instability at the fracture
site. Therefore, 14 and 12 animals were left for stati.stical
analyses in the 25-day and 40-day groups, respectively.
The mean (SD) weights at the end of the study of
animals in the 25-day and 40-day groups were 394.8 g
(39.7) and 417.7 g (37.3), respectively.

Effect of LIPUS on Fracture Site Radiographic Healing

Representative images of fractures in the 25-day and
40-day groups are shown in Figure 2. There were no
significant differences in radiographic scoring between
active LIPUS-treated fractures and inactive LIPUS-
treated fractures in either the 25-day group (P=.79) or
the 4()-day group (P=.26) (Table).

Effect of LIPUS on Fracture Site Bone Mass

There was no significant difierence in BMC between
active LIPUS-treated and inactive LIPUS-treated frac-
tures when assessed by DXA at 25 days (%diff=2.4%,
95% CI=-7.5%-12.5%) {P=.7l) (Fig. 3). Similarly,
BMC (%diff-4.1%; 95% CI=-6.3%-14.5%), vBMD
(%diff=0.4%, 95% CI=-9.1%-9.9%). and B.Ar
(%diff= -0.2%, 95% CI= -24.6%-24.3%) did not differ
between active LIPUS-treated and inactive LIPUS-
treated fractures at 25 days when assessed by pQCT (all
Ultimate force /'valnes^.81-.96) (Fig. 4). In contrast, at 40 days, active
LIPUS-treated fractures had 14.3% (95% CI=1.0%-
27.5%) greater fracture site BMC on DXA than inactive
LIPUS-treated fractures (P<.05) (Fig. 3). This increase
was confirmed by pQCT, which found BMC in active
LIPUS-treated fractures to be 16.9% (95% CI=2.3%-
31.4%) greater than that in inactive LIPUS-treated frac-
Energy to tures (P<.05) (Fig. 4A). The increase in fracture site
BMC with active LIPUS at 40 days did not result in an
ultimate force
increase in the amount of bone mineral per unit volume,
as vBMD did not differ from that in inactive LIPUS-
treated fractures (%diff=-4.7%, 95% CI=-12.4%-
0,2 0,4 0.6 2.9%) (P=.14) (Fig. 4B). Instead, there was an increase
in bone size, with active LIPUS-treated fractures ha\ing
Displacement (mm) 25.8% (95% CI=3.9%-47.6%) greater B.Ar than inac-
Figure 1. tive LIPUS-treated fractures (/*<.O5) (Fig. 4C).
(A) Experimental setup for testing of fracture site mecfianical properties
in 4-point bending, Tfie fracture site [wfiite arrow) was centered between
ttie upper contacts, (B) Representative force-displacement curve generated Effect of UPUS on Fracture Site Mechanical Properties
from a 4-point bending test of a fracture in tfie 40-day group, Mecfionical During mechanical testing, all femurs broke at the
properties derived from tfiis graph included ultimate force (peak of tfie curve fracture site. At 25 days, there were no significant
on tfie y-Gxis}, stiffness (slope of tfie linear portion of the curve), and energy
differences between active LIPUS-treated fractures and
to ultimate force [area under tfie curve to ultimate force).

1)22 . Warden etol Physical Therapy , Volume 86 , Number 8 , August 2006

 A

Inactive-LIPUS Active-LIPUS
treated treated

B D

Inactive-LIPUS Active-LlPUS
treated treated

Representative images of fractures at 25 ond 40 days postinjury, as obtained by radiography (A and B) and microcomputed tomography (C and D),
(A) Radiography at 25 days shows the persistence of a visible fracture line in fractures treated by both inactive low-intensity pulsed ultrosound (LIPUS)
and active LIPUS (white crrov/s), and associated callus formotion. Both fractures in this animal were radiographicolly scored as 1, (C) This score was
confirmed by microcomputed tomographic imaging, which clearly shows the fracture defect (block arrow) ond callus, (B) On radiography at 40 days,
tfie original frocture is difficult to distinguish [white arrows) because of callus bridging of the fracture gap. Both fractures in this animol were
rodiogrophically scored as 3, (D) This score was confirmed by microcomputed tomographic imoging, which shows fracture site union (block arrow).

inactive LIPUS-treated fractures in ultimate force
(%diff=2.6%, 95% CI= -41.2%-46.4%), stiffness
(%diff=4.4%, 95% CI=-77.3%-86.0%), or energ>' to
ultimate force (%diff-2.2%, 95% CI=-33.0%-37.3%)
(all /'values=.49-.66) (Fig. 5). In contrast, at 40 days-
active I.IPUS-treated fractures had 81.3% (95%
CI=0.8%-162.7%) greater ultimate force and 63.4%
(95% C( = 10.3%-llb.4%) greater stiffness than inactive
LIPUS-treated fractures (all Fvalues<.05) (Figs. 5A and
5B}. Compared with inactive LIPUS, active LIPUS had
no effect on energ>' to ultimate force at 40 days
(%diff= 146.3%, 95% CI = -37.8%-330.4%) iP=.lS)
(Fig. 5C). However, this latter finding most likely
resulted from insufficient statistical power to detect a
difference because of the variance within the data.
Discussion and Conclusions
The present study invcstigaled the effect of LIPUS
produced by a conventional therapeutic ultrasoiind unit
on fracture lepair in an animal model. LIPUS did not
have a significant effect on fracture healing when
assessed at 25 days postfracture. This finding may have
been influenced by insufficient statistical power, with
post hoc power analyses indicating that differences of
greater than 11% in side-by-side comparisons were
reqtiired in order to achieve 80% statistical power. In
contrast, by 40 days, fractures treated with active LIPUS
had significantly greater bone mass than fractures
treated with inactive LIPUS (placebo). This increase in
bone mass lesulted in an increase in hone size, as
opposed to an increase in bone density, and contributed
to active LIPUS-treated fractures having enhanced
mechanical properties compared with inactive LIPUS-
treated fractures. The latter was indicated by active
LIPUS-treated fractures having 81% greater ultimate
force and 63% greater stiffness than inactive LIPUS-
treated fractures. These data indicate that UPUS pro-
duced by a conventional therapeutic ultrasountl unit as
traditionally used by physical therapists may be used to
facilitate fracture repair.
The findings of this study are interesting from the
perspective that physical therapists traditionally have
been advised to avoid exposing the skeleton to excessive

Physicol Therapy . Volume 86 , Number 8 . August 2006 W o r d e n e t a l . 1123

 Table.
Effect of Active Low-Intensity Pulsed Ultrasound (LIPUS) and Inactive LIPUS on Radiographic Scaring of Fracture Healing

No. of Animals With the Following

Days Radiographic Score"
(No. of Animals) LIPUS 0 1 2 3 X SD
25(14) Inactive 1 11 1 1 1,1 0,7
Active 1 10 2 1 1.2 0.7
40(12) Inacfive 0 6 2 4 ).8 0,9
Active 0 4 3 5 2,1 0,9
' 0 - n o .-videiicf al healing. 1 -ralliis Ioiiimtioii bui iraniuT gap nol yti bridgt^d. i;=tallLi,s formation t-ddfin wiili possible uf the tiacii.re L'tu
3=fracture union.

Inactive-LIPUS
ment head for 20 minutes a day for 6 weeks. In contrast,
Active-LIPUS similarly introduced ultrasound at a lower intensity
0.20 - (0.5 W/cm^) had no adverse effect on bone growth.

The data of the present study support tbe results of

0.15 - previous animal studies'^-"^'^ and clinical studies*'-^ dem-
onstradng that LIPUS accelerates fracture repair, and
furthers this research by demonstrating that LIPUS
O 0.10 - produced by a conventional therapeutic ultrasound unit
as used by physical therapists may be used to facilitate
0.05 - fracture repair. This latter finding addresses an area of
recent contention.io it was previously suggested that
LIPUS produced by conventional therapeutic ultra-
0.00 sound units may delay fracture healing by stimulating
25 days 40 days the production of excessive nonmineralized cartilage.-'
However, the data of the pre.sent study challenge tbis
Figure 3. hypothesis. First, we found that fractures treated with
Effect of lovi/-intensity pulsed ultrasound [LIPUS) on fracture site bone active LIPUS achieved tbe same level of radiograpbic
mineral content (BMC), as assessed by dual-energy X ray absorptiom-
etry. Bars represent mean:!:SD, An asterisk indicates data that were
healing and had more mineralized callus formation
significantly different from those for inactive LIPUS-treoted fractures {greater fracture site bone mass) than inactive LIPUS-
(P<.05, paired ftest). treated fractures. Second, fractures treated with active
LIPUS had better fracture site mechanical properties
than inactive LIPUS-treated fractures. The restoration of
amounts of ultrasoimd energy. Reflecting this fact, only mechanical integrity is tbe overall function of any repair
1% of therapists currently introduce ultrasound energy process in a load-bearing structure such as bone. There-
with the intent of treating acute bone injuries.' However, fore, we believe that LIPUS produced by a conventional
this dogma is historically hased and does not incorporate tberapeutic ultrasound unit can be beneficial to the
current research findings. There is no douht that ultra- fracture repair process and does not delay bone union.
sound energy can produce significant tissue damage
when applied to the skeleton because of unique biophys-
Although a significant beneficial effect was observed in
ical interactions between ultrasound and bone. This fact
the present study, therapists are not currently encour-
has been confirmed experimentally by ultrasound caus-
aged to introduce LIPUS produced by their conven-
ing premature closure, slipping, and displacement of
tional therapeutic ultrasound units with the intent of
epiphyseal growth plates, bone sclerosis, diaphyseal frac-
facilitating clinical fracture repair. Animal studies are
tures and fibrosis, and delayed healing during fracture
necessary precursors in the initial investigation of the
repair.' '-'^ However, these effects have been elicited only
safety and efficacy of an intervention; however, tbe
by ultrasound doses at the high end of the therapeutic
examination of interventions in laboratory-based studies
dose range (>1.0 W/cm"). To date, there are no known
does not always accurately predict clinical effects. This
side effects of LIPUS application (<0.1 W/cm^) on the
fact is particularly pertinent to ultrasound therapy stud-
skeleton.^'«"* Supporting this fact, a recent study^"
ies, as the relative size, volume, and depth of tbe tissue
demonstrated that pulsed ultrasound therapy at an
being treated in animals typically differ from tbose of
intensity of 2.2 W/cm^ produced pathologic changes in
tissue being treated clinically. Tbese differences may
growing bone when introduced with a stationary treat-

1 24 , Warden et al Physical Therapy , Valume 86 , Number 8 . August 2006

 A
240 1 Inactive-LIPUS 2 4 0 -1 • • Inactive-LIPUS •
Active-LIPUS ^ H Active-LtPUS
T
^ 180 -
o
o
^ 120 -
I II
B
"to
1 60 - T T ••_
0 -
25 days 40 days 25 days 40 (Jays
*
B B ' • Inactive-LIPUS
1000 n inactive-LIPUS '-' 25 - T
m^ Active-LIPUS
Active-LIPUS
f 20 -
I
o
15 -
10 -
11 •

II
•
c

• s•
iff

CO 5 -

0 -
25 days 40 days
25 days 40 days
c - "3 125 n • • Inactive-LIPUS
36 -| Inactve-LIPUS ^ B Active-LIPUS • •

Active-LIPUS
8 100-
o
27 -
% 75 -
1
18 - 1
o
50- 1
CD r, 25 -
9-
m
erg:

0 ^
25 days 40 days
Figure 4.
Effect of low-intensity pulsed ultrasound (LIPUS) on fracture site bone
mineral content (BMC) (A), volumetric bone minerol density (vBMD) (B),
and bone area (B.Ar) (C) as assessed by peripheral quantitative
computed tomography, Bors represent mean ± SD. An asterisk indicates
data that were significantly different from those for inactive LIPUS-treated
Fractures (P<-05, paired Mest),
25 days 40 days
Figure 5.
Effect of low-intensity pulsed ultrasound (LtPUS) on fracture site mechan-
ical properties. Ultimate force (A), stiffness (B), and energy to ultimate
force (C) were assessed by destructive 4-point bending tests. Bars
represent meon ± SD, An asterisk indicates data that were significantly
different from those for inactive LIRUS-treated fractures (P<,05, paired f
test).

influence ultrasound energy distributions, tissue interac-
tions, and ultimately therapeutic responses. In order for
the results of the present study to have clinical relevance,
the observed LIPUS effect needs to be confirmed by way
of controlled clinical trials. In addition, before LIPUS
intervention can be contemplated clinically, the ongoing
concern regarding the output performance of ultra-
sound units being used in clinical practice needs to be
addressed. Equipment surveys undertaken globally
repeatedly have found that many ultrasound units being
used in clinical practice are unable to produce an
ultrasound dose that matches the metered dose to within
set standards.'-^'-^'-^'' This output variance may not only
influence treatment efficacy during fracture repair but
also elicit detrimental effects. Until these current limita-
tions are addressed, the use of conventional therapeutic

Physical Theropy , Volume 86 , Number 8 , August 2006 Warden et al , 1 1 25

 ultrasound units in a manner other than that approved
by US Food and Drug Administration market compli-
ance could have potential ramifications.
Although the present data confirm that LIPUS facilitates
fracture repair, they do not contribute to the current,
limited understanding of the mechanism underlying this
effect. Considering that LIPUS introduces an intensity
within a more traditional diagnostic ultrasound range, a
range previously considered to have a minimal biologic
effect and no therapeutic value, it is valid to consider
how LIPUS induces its therapeutic eflect. Unfortunately,
this mechanism is not yet known, as it is not established
how ultrasound signals are transduced in vivo to produce
a cellular response. It is possible that ultrasound longi-
tudinal mechanical waves exert micromcclianical load-
ing to manipulate the inherent mechanosensitivity of
bone cells. However, this notion has been disputed by
studies demonstrating that the mechanical loading a,s,so-
ciatcd with LIPUS does not induce adaptation in intact
bone.!'''-'-* Alternatively, the beneficial effect of LIPUS
during fracture repair may result from the generation by
ultrasound of unique phenomena within the propagat-
ing tLssiies, such as stable cavitation and microstreaming.
These phenomena may generate sliear forces on cellular
membranes to induce a cellular response; however, the
occurrence and significance of these phenomena in vivo
have been disputed.--' Finally, LIPUS may have iLs bene-
ficial effect during fracture repair via the generation of
localized heat at the fracture site in response to molec-
ular vibration and collisions. However, this mechanism
lacks the support of a recent study,^^ which found that
ultrasound therapy augmented fracture repair but that
an equivalent level of hyperthennia generated by micro-
wave therapy did not.
Despite the fact that the underlying biophysical mecha-
nism of action of ultrasound during fracture repair is not
known, a number of studies have investigated potential
cellular processes influenced by LIPUS. hi vitro, LIPUS
has been shown to influence direcdy a number of cells
associated with the repair process, including fibro-
blasts/-''-^^ chondrocytt's,29-'*' and osteoblasts.'^-'" The
induced changes suggest that ultrasound may have a
direct effect on the reparative processes of angiogenesis,
chondrogenesis, and osteogenesis. This suggestion is
supported by in vivo investigations.'^"'^'' Principally,
Azuma et aV-' showed that LIPUS influenced multiple
cellular reactions during fracture repair. This finding
was evident from the advancement of healing irrespec-
tive of the phase of repair during which LIPUS was
introduced.
In summary, the present study showed that LIPUS
produced by a conventional therapeutic ultrasound unit
can facilitate fracture repair. This finding was evident by
1 1 26 . Warder ef al
active LIPUS-treated fractures having greater fracture
site bone mass, size, and strength than within-animal
inactive LIPUS-treated fractures. These data provide
preliminary evidence to support a beneficial effect of
LIPUS as produced by an ultrasound unit traditionally
used by physical therapists on fracture repair. However,
careful interpretation of this controlled laboratory study
is warranted until its findings are confirmed by clinical
trials. Until these trials are performed and until the
accurate output performance of their ultrasound luiits is
ensured, therapists are not encouraged to introduce
LIPUS produced by a conventional therapeutic ultra-
sound unit with the intent of facilitating clinical fracture
repair.
References
1 W:irtk-ii S), McMceken JM, UltrasoLind usage and dosage in spom
physiotherapy. Ultrasound Med Bwl. 2
2 Mitragolri S, Healing sound: the use of ultrasound in drug deliveiT
and other iherapetitic applications. Nat Rni Dntg Discov. 2(K)5;4:
3 Rubin C:, Bolander M, Ryahy ]P, Hadjiargyroii M, The use of
low-inirnsity pulsed ultiasouiid to acctfk-rate the healing of tracuues.
I Hone Joint Surfr Am. 2(101:83:239-270.
4 Warden SJ. Bennell KL, McMeeken JM, Wark JD. Accek-rati.jn ol
(resh fracture repair using ilie Sonic Acceleraied Fraciure Hi-aling
Sy.sltMn (SAFHS): a review. Cakif Tissue Int. 2()0();(i6:157-163.
5 Pilla .W, Mont MA. Nasser PR. etal. Non-invasive low-intensity
pulsed ultrasound arcelerates hone healing in the rabbit, / (hlhop
Trmima. 1990;4:24(i-2,5.S.
6 Heckmati JD, Ryaby JP. McCabe J. et al. Acceleration of tibial
Iratlure healing by non-invasive, low-intensit)' pulsed ultrasound. /BOTJP
JmntSiirgAm. 1994:70:26-34.
7 Kjistiiuiseii TK, Ryaby J P. MtCabe J. etal. Accelerated healing of
distal radiits Iracuires with the use of specific, low-intensity ultrasound.
/ Bone Joint .Surg Am. 1997:79:961-973.
8 Mayr E. Rutzki MM. Rut/ki M, et al, BeschU-unlgt niedrig intensiver.
gepulster ntra.schall die Heihuig von Skaphoidfrakturen? fhmdrhn
Mikrochir 1'la.tt Chir. 20(K);32:i 1,5-122.
9 litLsse JW, Bhandari M. Kulkanii AV, Tunks E. The eOect of low-
intensity pulsed ultrasound therapy on time to fracture healing: a
meta-analysis. Can Med AssocJ. 2002:166:437-441.
10 Warden SJ. Bennell KL. McMeeken JM. WarkJD. Can conventional
therapeutic ultrasnund units be used to accelerate fracture repair? I'liys
Ther HFIK 1999:4:117-126,
!1 Ardan NI, Janes JM, HerrickJF. Ultrasonic energ) and surgically
produced defects in hone. J Hone Joint Surg Am. 1957;39:394-402,
12 DeForest RE. Herrick JF. Janes JM. Krusen FH. EfTecLs of ultra.sound
on growing bone. Arch Phys Med Rehabit. 1953;34:21-30.
13 Warden Sj. A new direction for ultrasound tberapy In sports
medicine. Sports Med. 2003;33:95-107.
14 Ziskin MC, Report on the safety of the Therasonics Medical Systems
SAFHS Unit, Model 2A. In: PMA900009 19H9;3(section V l . A l ) :
209-234.
Physical Therapy , Volume 86 , Number 8 , August 2006
15 Azuma Y. h o M, Harada Y, et al. Low-intensity pulsed ultrasound
accelerates rat femora] fracture healing by acting on the various
cellular reactions in the fracture callus. / Bone Miner Res. 2001 ;16:
671-680.
16 Rawno! NM. f-oldberg BB. Forsberg F. et al. Power Doppler assess-
ment of \-asctilar changes during fracture treatment with low-int,ensit>'
uhrasound. J Ultrasound Med. 2003;22:145-153.
17 Wang S:|. Lewallen DG. Bolander ME. etal. Low intensitv- ultra-
sound treatment increases strength in a rat femoral fracture model.
JOnhof)R>'s. 1994:12:40-47,
18 Spadaro JA, Albanese SA, Application of low-intensity tiltrasound to
growing bone in rats. Ultrasound Med Biol. !998;24:567-573.
19 Warden SJ, Bennell KI.. Forwood MR. et ai. Skeletal effects of
low-intensit> pulsed ultrasound on the ovariectomized rodent. Ullra-
sottnd Med Biol. 2001;27:989-998.
20 Lvon R. Liu XC. Meier J. The effects of therapeutic vs, high-intensity
ultrasound on the rabbit growth plate, J Ortkop Res. 2003;21:865-871.
21 Dyson M, Brookes M. Stimulation of bone repair by ultrasound.
Ultrasound Med Biol. 1983;2(suppl):61-66.
22 Artho PA. Thyne JG. Warring BP, et al, A calibration study of
therapeutic ultrasound units. Phys Ther. 2002;82:257-263,
23 Hekkenbei^ RT. Beissner K, Zeqiri B, et al. Validated ultrasonic power
niea.stirements up to 20 W, Ultrasound Med Biol. 2001:27:427-438.
24 Warden SJ, Bennell KI-, Matthews B. et al. Efficacy of low-intensity
pulsed ultrasound in the prevention of osteoporosis of the calcaneum
following spinal cord injury. Bone. 2001;27:431-436.
25 Baker KG, Robertson W], Duck FA. A review of therapetitic ultra-
sound: biophysical etTects, Phys Ther. 2001:81:1351-1358.
26 (.hang WH-S. Sun J-S. Chang S-P, Lin JC:, Study of the thermal
effects of ultrasound stimulation on fracture healing. Bioekctromagnet-
ics. 2002:23:256-263.
Physical Therapy , Volume 86 , Number 8 , August 2006
27 Doan N, ReberP. Megliji S, Harris M. hi vitro effects of therapeutic
ultrasound on cell proliferation, protein synthesis, and cytokine pro-
duction by human fibroblasts. osteoblasts, and monocytes. / Oral
Maxillofac 'surg. 1999:57:409-419.
28 Zbou S, Schmelz A. Seutterlein T. et al. Molecular mechanisms of
low intensit)' pulsed ultrasound in htiman skin fibroblasts. / Bzo/ Chem.
2004 ;2 79:54463-54469,
29 Parvizi J. Parpura V. Greenleaf JF, Bolander MF. Calcium signaling
is required for ultrasound-stimulated aggrecan syntbesis by rat chon-
dmcytcs. J Orthop Res. 2002:20:51-57.
SO Parvizi J. Wu C-C, Lewallen DG, et ai. Low-intensity ultrasound
stimulates proteogiycan syntbesis in rat chondrocytes by increasing
aggrecan gene expression, J Orthop Res. 1999; 17:488-494.
31 Zhang ZJ, Huckle J, Francomano CA. Spencer RG. The effects of
pulsed low-intensity ultra.sound on cbondrocyte viability, proliferation.
gene expression and matrix production. Ultrasound Med Biol. 200S;29:
1645-1651.
32 Naruse K, Mikuni-Takagaki Y, Azuma Y. et al. Anabolic response of
mouse bone-niarrow-derived stromal cell clonal ST2 cells to low-
intensity pulsed ultrasound. Biochem Biophys Res Commun. 2000;268:
216-220.
33 Sena K, Leven RM. Mazbar K, etal. Early gene response to
low-intensity pulsed ultrasound in rat osteoblastic cells. Ultrasound Med
Biol. 2005:31:703-708.
34Warden SJ. Favaloro J. Benneil KL.. etal. Low-intensity pulsed
ultrasound stimulates a bone-forming response in UMR-106 cells.
Biorhem Biophys Res Commun. 2001;286:443-450.
35 Vang K-H, Parvizi J. Wang SJ. etal. Exposure to Iow-intensity
ultrasound increases aggrecan gene expression in a rat femur fracture
model. J Onhop Res. 1996;14:802-809.
Warden et al . 11 27