Access to this full-text is provided by Wiley.
Content available from BioMed Research International
This content is subject to copyright. Terms and conditions apply.
Hindawi Publishing Corporation
BioMed Research International
Volume , Article ID , pages
http://dx.doi.org/.//
Research Article
Pharmacodynamic Analysis of Magnetic Resonance
Imaging-Monitored Focused Ultrasound-Induced Blood-Brain
Barrier Opening for Drug Delivery to Brain Tumors
Po-Chun Chu,1Wen-Yen Chai,1,2 Han-Yi Hsieh,1Jiun-Jie Wang,3Shiaw-Pyng Wey,3
Chiung-Yin Huang,4Kuo-Chen Wei,4and Hao-Li Liu1
1DepartmentofElectricalEngineering,Chang-GungUniversity,259Wen-Hwa1stRoad,Kwei-Shan,Tao-Yuan333,Taiwan
2Department of Diagnostic Radiology, Chang-Gung University and Memorial Hospital, 5 Fu-shin Street, Kwei-Shan,
Tao-Yuan 333, Taiwan
3Department of Medical Imaging and Radiological Sciences, Chang-Gung University, 259 Wen-Hwa 1st Road, Kwei-Shan,
Tao-Yuan 333, Taiwan
4Department of Neurosurgery, Chang-Gung Memorial Hospital, 5 Fu-shin Street, Kwei-Shan, Tao-Yuan 333, Taiwan
Correspondence should be addressed to Hao-Li Liu; haoliliu@mail.cgu.edu.tw
Received January ; Accepted February
Academic Editor: Fan-Lin Kong
Copyright © Po-Chun Chu et al. is is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Microbubble-enhanced focused ultrasound (FUS) can enhance the delivery of therapeutic agents into the brain for brain tumor
treatment. e purpose of this study was to investigate the inuence of brain tumor conditions on the distribution and dynamics
of small molecule leakage into targeted regions of the brain aer FUS-BBB opening. A total of animals were used, and the
process was monitored by T-MRI. Evans blue (EB) dye as well as Gd-DTPA served as small molecule substitutes for evaluation of
drug behavior. EB was quantied spectrophotometrically. Spin-spin (R1) relaxometry and area under curve (AUC) were measured
by MRI to quantify Gd-DTPA. We found that FUS-BBB opening provided a more signicant increase in permeability with small
tumors. In contrast, accumulation was much higher in large tumors, independent of FUS. e AUC values of Gd-DTPA were well
correlated with EB delivery, suggesting that Gd-DTPA was a good indicator of total small-molecule accumulation in the target
region. e peripheral regions of large tumors exhibited similar dynamics of small-molecule leakage aer FUS-BBB opening as
small tumors, suggesting that FUS-BBB opening may have the most signicant permeability-enhancing eect on tumor peripheral.
is study provides useful information toward designing an optimized FUS-BBB opening strategy to deliver small-molecule
therapeutic agents into brain tumors.
1. Introduction
Focused ultrasound beams (FUS) in the presence of cir-
culating microbubbles can temporarily open the blood-
brain barrier (BBB opening) of capillaries in the central
nervous system (CNS) parenchyma [–]. Bursts of acoustic
ultrasound induce microbubble cavitation in the vasculature,
and the resultant shear stress temporarily disrupts tight
junctions to enhance blood-brain permeability. is BBB-
opening process can be carried out at moderate acoustic
pressures to minimize adverse eects on vascularture and
prevent damage to neurons [–], while facilitating localized
delivery of chemotherapeutic agents from the vasculature to
the pathological brain parenchyma and CNS [–]. Since
more than % of the therapeutic agent normally cannot pen-
etrate CNS tight junctions [],thisnovelapproachprovides
a unique opportunity for local delivery of therapeutic agents
across the BBB and into the targeted site, thus opening a new
frontier of CNS drug delivery.
Brain tumors could potentially be treated by FUS-BBB
opening to enhance chemotherapeutic agent delivery. In
the United States, at least , patients are diagnosed
with glioblastoma multiforme (GBM) each year, comprising
more than half of the malignant primary brain tumors [],
BioMed Research International
Function
generator
Power
amplier
Power
meter
Targeted
position
Animal
FUS
transducer
(400 kHz)
Ultrasound
focal beam
Disgassed
water
F : Schematic showing the experimental setup of the focused ultrasound exposure system.
and chemotherapy is an important treatment modality [].
Recent preclinical studies showed that FUS-BBB opening
can eectively enhance local deposition and concentration of
chemotherapeutics including BCNU [], liposomal doxoru-
bicin [,], and chemodrugs carried by novel nanocarriers
[].
Currently, FUS-mediated CNS drug delivery is moni-
tored by magnetic resonance imaging (MRI) by intravenous
(IV) injection of gadolinium diethylenetriamine penta-acetic
acid (Gd-DTPA) contrast agent together with chemother-
apeutic drug. e MRI signal intensity increase caused by
the leakage of Gd-DTPA thus serves as an indicator to
estimate drug concentration [,,,]. Most studies have
focused on analyzing the eects of FUS exposure parameters
such as acoustic pressure amplitude, ultrasound frequency,
pulse length, pulse repetition frequency, exposure duration,
and microbubble dose on BBB opening [,–]. How-
ever, pharmacodynamic analysis including the dynamics and
distribution of the specic molecular agent in the brain
is critical for evaluating specic drug delivery. MRI could
also be used for pharmacological endpoint evaluation using
concurrently administered MR contrast agents as surrogate
indicators of therapeutic drug concentrations. Although the
kinetics of contrast agent permeability of a defective blood-
brain barrier have been measured using MR compartment
modeling [–], these studies were performed in normal
animal brains, and so far the detailed pharmacodynamic
behavior of contrast agents aer FUS-BBB opening remains
uncertain in brain tumors.
e purpose of this study was to conduct an MRI
pharmacodynamic analysis of FUS-BBB opening in brain
tumors in an animal model. Injected Gd-DTPA contrast
agent was used to characterize pharmacodynamic changes
as a function of time aer BBB opening, in both normal
andbrain-gliomaanimals.Wealsoattemptedtoestablish
the correlation between deposition of Gd-DTPA by in vivo
semiquantication and the quantitation of another surrogate,
Evans blue dye, by spectrophotometry aer sacrice. Finally,
we evaluated the pharmacodynamic changes aected by FUS-
BBB opening in various grades of gliomas.
2. Methods
2.1. FUS Setup. A focused ultrasound transducer was used
to generate ultrasound focal energy (IMASONIC, France;
diameter = mm, radius of curvature = mm, frequency =
kHz, and electric-to-acoustic eciency = %) (Figure ).
An arbitrary function generator (A, Agilent, Palo Alto,
CA and DS, Stanford Research Systems, Sunnyvale, CA)
was used to generate the driving signal, which was then fed
into a radiofrequency power amplier (AB, Amplier
Research, Souderton, PA). e focal zone distribution of
the intensity of the ultrasound eld was measured in an
acrylic water tank lled with deionized, degassed water. e
measured diameter of the half-maximum pressure amplitude
wasmm,andthelengthoftheproducedfocalzonewas
mm. Animals underwent isouorane anesthesia before
ultrasound treatment. e animal was laid prone and placed
directly under an acrylic water tank (with a window of
×cm
2at its bottom sealed with a thin lm to allow
entry of the ultrasound energy), using ultrasound gel to
ll the interspaces between the animal head and the thin-
lm window. SonoVue SF-coated ultrasound microbub-
bles (– 𝜇mmeandiameter,.𝜇L/kg; Bracco Diagnos-
tics Inc.) were IV administered by burst injection with
. mL of saline solution containing .mL heparin. Aer
injecting the microbubbles, burst-tone mode ultrasound at
a pressure of . MPa (peak negative value; measured in
the free-eld) was delivered to the brain with the center
of the focal zone positioned at a penetration depth of -
mm under the scalp (burst length = ms, pulse repe-
tition frequency = Hz, and total sonication duration =
s).
BioMed Research International
I: before ∗∗ : group 2
animals only
Sacrice
Implant FUS
Intervention
Gd Gd Gd∗∗
IV inject
MRI
10 or 17 d 4 h
T
(a) (b)
(c)
1400
700
0
MBs
and EB
80 min
10 min 20 min 40 min 80 min
T1/R1/T2T1/R1/T2T1/R1/T∗∗
2
II: aer,
10 min
III: aer,
2 h∗∗
F : (a) Experimental protocol for st- and nd-group animal experiments. MRI images were acquired in time slots I/II for group
animals and in time slots I/II/III for group animals; double asterisks (∗∗) indicate group- experiments only. (b) FUS-exposed brain
of a normal animal with Evans blue extravasations to identify the location of BBB opening. (c) R1accumulating map showing Gd-DTPA
accumulation in the BBB-opening location of a normal animal over time.
2.2. Animal Experiment Design. All animal experiments
were approved by the Institutional Animal Care and Use
CommitteeofChangGungUniversityandadheredtothe
experimental animal care guidelines. A total of animals
(male Sprague-Dawley rats (– g)) were used, including
normal (𝑛=18) and tumor animals (𝑛=16). Experi-
ments were divided into two groups. In group , the aim
was to conrm the correlation between Gd-DTPA leakage
(concentration measured by relaxometry) and Evans blue
(EB) dye (concentration measured spectrophotometrically)
aer FUS-BBB treatment. Subgroups included () normal rats
(𝑛=18) and () tumor rats (𝑛=4), and the rst subgroup
underwent FUS-BBB opening. Subgroups were conrmed
by dynamic contrast-enhanced (DCE) MRI with Gd-DTPA
(molecular weight = Da). In addition, EB dye (molecular
weight = Da) was IV injected into the animals, and
the amount of EB deposited in the brain was quantied
spectrophotometrically (procedure described below). In the
rst subgroup of group , contrast-enhanced T1-weighted
imaging was rst performed to estimate Gd-DTPA concen-
tration aer BBB opening, followed by T2-weighted imaging
to provide a reference of tumor morphology. e second
subgroup underwent the same scanning process without FUS
induced.
In experimental group , our aim was to monitor the
increase in Gd-DTPA accumulation in tumor-bearing ani-
mals aer conducting FUS-BBB opening. Animals were
divided into two subgroups: () animals receiving FUS
exposure days aer tumor implantation (tumor volume
typically <. cm3) with FUS-BBB opening (𝑛=6)and
() animals receiving FUS exposure days aer tumor
implantation (tumor volume typically >. cm3) with FUS-
BBB opening (𝑛=6). Tumor volume was measured by
T2-weighted MRI. In group , animals were subjected to
three -minute-long MR relaxometry-based imaging ses-
sions (before FUS exposure, and -min and -min aer
FUS exposure). Detailed experimental procedures are shown
in Figure .
2.3. Rat Brain Glioma Model. C glioma cells were harvested
by trypsinization and cultured at a concentration of ×
5cells/mL for implantation. For intracranial injection into
the striatum of rat brains, cells were washed once with
phosphate buered saline (PBS). Male Sprague-Dawley rats
(– g) were anesthetized by intraperitoneal admin-
istration of ketamine (mg/kg) and immobilized on a
stereotactic frame. A sagittal incision was made through the
BioMed Research International
skin overlying the calvarium, and a small dental drill was used
to make a hole in the exposed cranium, . mm anterior and
mmlateraltothebregmaonthelesideofskull.Ccell
suspension ( mL) was injected at a depth of . mm from
the brain surface. e injection was performed over a -
minute period, and the needle was withdrawn over another
minutes. Ten days aer implantation, tumor sizes were
measured by MRI.
2.4. Spectrophotometric Quantitation of Evans Blue Dye. EB
dye (% in saline) was IV injected ( mg/kg), and the
animalsweresacricedtwohourslater.Allanimalswererst
deeply anesthetized with % chloral hydrate and infused
with heparinized saline through the cardiac ventricle until
colorless infusion uid was obtained from the atrium. Aer
the rats had been sacriced by decapitation, the hemispheres
of the brain were separated along the transverse suture. en
both hemispheres were weighed and placed in formamide
( mL/ mg) at ∘C for h. e sample was centrifuged
for mins at , rpm. e concentration of dye extracted
from each brain was determined spectrophotometrically at
nm and was compared with a standard graph created
by recording optical densities from serial dilutions of EB in
.% sodium chloride solution. e EB tissue content was
quantied using a linear regression standard curve derived
from seven concentrations of the dye.
2.5. MRI. For in vitro measurements, Gd-DTPA (Omniscan,
. mL/kg, Magnevist) was diluted with physiological saline
to ., ., ., ., ., and . 𝜇M. Circular wells (inner
diameter = mm) were lled with 𝜆of contrast agent
sample or physiological saline as control and were placed
in the MR scanner (Clinscan, Bruker, Germany; Tesla).
Spin-lattice relaxivity maps were calculated from two T1-
weighted images with dierent ip angles (gradient recalled
echo sequence, TR/TE = . ms/. ms, slide thickness =
. mm, matrix = ×, and ip angle = ∘/∘). e
correlation between R1(= /T1)mappingandGd-DTPA
concentration was determined [].
In the animal experimental group, FUS-induced BBB
opening was monitored by MRI with a -Tesla magnetic reso-
nance scanner (Bruker ClinScan, Germany) and a -channel
surface coil. e mouse was placed in an acrylic holder,
positioned in the center of the magnet, and anesthetized with
isourane gas (-%) at – breaths/min during the entire
MRI procedure.
In the rst experimental group, the distribution and
dynamics of Gd-DTPA leakage were investigated imme-
diately aer conducting FUS-BBB opening. Aer FUS-
BBB opening, animals were immediately relocated into the
MR scanning room, and contrast-enhanced T1-weighted
images with dierent ip angles were acquired to calcu-
late spin-lattice relaxivity maps by transferring two images
with dierent ip angles (gradient recalled echo sequence,
TR/TE = . ms/. ms, slice thickness = . mm, slice
number=,matrix=×, and ip angle = ∘/∘).
Images were sequentially acquired over min with a time
interval of seconds for area under the curve (AUC)
calculation. Upon completion of the th acquisition, a
diluted bolus of Gd-DTPA was IV injected through a
catheter at an infusion rate of mL/s. In the second
experimental group, three sets of Gd-DTPA-leakage dis-
tribution/dynamics were investigated, including (I) before
FUS exposure, (II) immediately aer FUS exposure, and
(III) two hours aer FUS exposure. Immediately aer
conducting FUS-BBB opening, turbo spin echo (TSE) T2-
weighted images were obtained as a reference to identify
the tumor region (repetition time (TR)/echo time (TE) =
/ ms, FOV = × mm2,in-planeresolution=
. ×. mm2, and slice thickness = . mm).
2.6. MR Analysis of Gd-DTPA Accumulation and Distribution
aer FUS-BBB Opening. In R1-map analysis, a region of
interest (ROI) was selected and compared with the non-
enhanced contralateral brain to determine the increase in
Gd-DTPA concentration caused by BBB opening. AUC maps
were then transferred from a series of time-dependent R1
maps (up to min) to determine pharmacodynamic char-
acteristics of Gd-DTPA for comparison with the dynamics of
EB dye permeability. us, the total area (AUC) is given by
the following equation:
AUC80 min =∫𝐶𝑝𝑡⋅𝑑𝑡
𝑉,()
where 𝐶𝑝𝑡 are vertical segments under the Gd-DTPA con-
centration curve area and 𝑉is total ROI volume.
In experimental group , ROIs were selected in the tar-
geted tumor area which was based on the tumor dimensions
denedinT
2images (the same ROI as in the contralateral
brain was selected). e distribution and dynamics of Gd-
DTPA leakage were evaluated for dierent tumor sizes
including days aer implantation (typically <. cm3)
and days aer implantation (typically >. cm3)andwere
divided by the tumor dimension. ROI including the entire
tumor and the contralateral area were selected. Moreover,
in order to evaluate the homogeneity of Gd-DTPA leakage,
tumors with dimension >. cm3were further divided into
tumor core (inner half of the area) and tumor peripheral
(outer half of the area), based on T2images.
2.7. Histology. Albumin-bound EB dye was IV injected as
a bolus immediately aer sonication. BBB opening was
quantied as extravasation of EB. Tumor model animals
were sacriced about h aer sonication and MR scanning.
Brain samples were serially sectioned ( 𝜇mthickness)using
the same slice direction as in MRI analysis. Representative
sections were stained with hematoxylin and eosin (HE).
Tumor morphology was histologically evaluated.
3. Results
BBB opening was clearly evidenced by staining with EB
dye. A typical image of a normal BBB-opened brain stained
with EB dye is shown in Figure (b). In addition, a series
BioMed Research International
100
80
60
40
20
0
1
2
3
4
5
6
7
8
(a)
20
15
10
5
0
−5 012345
Slope =3.2936
𝑦=3.2936𝑥 − 0.1446
𝑅2= 0.9991
Gd-DTPA (𝜇M)
R1(s−1)
(b)
0.25
0.2
0.15
0.1
0.05
000.2 0.4 0.6 0.8 1
Evans blue (𝜇g)
𝑅2= 0.9992
OD (620 nm)
𝑦=0.1945𝑥 + 0.0353
(c)
F : (a) T1image and corresponding R1map for the in vitro Gd-DTPA phantom at increasing concentrations ( and : water; : . 𝜇M;
: . 𝜇M; : . 𝜇M; : . 𝜇M; : . 𝜇M; : . 𝜇M). (b) Dependence of R1on Gd-DTPA concentration; relaxivity was estimated as about
. s−1 mM−1. (c) Calibration curve of spectrophotometrically determined EB concentration. EB concentrations ranging from to .𝜇g
were tested, resulting in O.D. readings of . to . at nm.
R1
Before FUS Aer FUS Before FUS Aer FUS Before FUS Aer FUS
NormalDay-10Day-17
1400
700
0
1400
700
0
1400
700
0
2
1
0
2
1
0
2
1
0
(a) (b) (c) (d)
AUC80 min
CE-T1T2
F : Typical MRI for normal animals (upper) as well as animals aer -day (middle) and -day (bottom) tumor implantations. (a)
Contrast-enhanced T1images before and aer FUS exposure. (b) T2images (aer FUS). (c) R1maps before and aer FUS exposure. (d) Area
under the R1curve over minutes (denoted as AUC80 min) before and aer FUS exposure.
of R1maps obtained at dierent time points aer FUS-
BBB opening demonstrated the dynamic change in Gd-
DTPA accumulation in a normal brain, with particularly high
leakage at the sonication site (Figure (c)).
R1relaxivity of Gd-DTPA and ELISA measurements of
EB dye concentration were calibrated in vitro. e detected
R1-signal increased in a highly linear manner with Gd-DTPA
concentration (input concentrations of , ., ., , , and
𝜇M) as shown by the calibration curve (𝑟2= 0.9991)
(Figure (b)). e relaxivity of Gd-DTPA contrast agent was
found to be . at Tesla. e detected ELISA signal also
increased in a highly linear manner with EB concentration
(Figure (c);𝑟2= 0.9992). ese calibration curves thus
allowed precise quantitation of Gd-DTPA and EB deposition
in the brain.
FUS-induced BBB opening was veried by CE-MRI.
Typ i cal CE-T1images, T2images, R1maps, and AUC maps in
normal, -day glioma, and -day glioma animals are shown
in Figure . In normal animals, the BBB-opened area was
clearly visible in T1-weighted images. T2-weighted images did
not show any evidence of FUS-induced damage at the target
location at pressure amplitudes of . MPa (Figure ). In the
BioMed Research International
Normal
2
1
0
010 20 40 60 80
0.8
0.6
0.4
0.2
0
Time (min)
Concentration (𝜇M)
R1(s−1)
Target, w/o FUS
Target, with FUS
Contralateral
(a)
2
1
0
010 20 40 60 80
0.8
0.6
0.4
0.2
0
Time (min)
Concentration (𝜇M)
R1(s−1)
Contralateral
Tumor, 10 d
Tumor, w/o FUS
Tumor, with FUS
(b)
2
1
0
010 20 40 60 80
0.8
0.6
0.4
0.2
0
Time (min)
Concentration (𝜇M)
R1(s−1)
Contralateral
Tumor, 17 d
Tumor, w/o FUS
Tumor, with FUS
(c)
Target, w/o FUS
Target, with FUS
Contralateral
5500
3500
1500
0
1200
800
400
0
R1accumulation
Gd accumulation (𝜇M)
010 20 40 60 80
Time (min)
Normal
(d)
Contralateral
5500
3500
1500
0
1200
800
400
0
R1accumulation
Gd accumulation (𝜇M)
010 20 40 60 80
Time (min)
Tumor, 10 d
Tumor, w/o FUS
Tumor, with FUS
(e)
Contralateral
5500
3500
1500
0
1200
800
400
0
R1accumulation
Gd accumulation (𝜇M)
010 20 40 60 80
Time (min)
Tumor, 17 d
Tumor, w/o FUS
Tumor, with FUS
(f)
F : (a)–(c) R1change as a function of time before and aer FUS exposure for normal animals, -day and -day tumor animals. (d–f)
e corresponding R1accumulation as a function of time over minutes for (a)–(c).
rst subgroup of experimental group , the R1-map signal of
the BBB-opening area was increased from . to about . by
FUS, and AUC maps showed an increase in accumulation of
Gd-DTPA deposition from to about .
In tumor-bearing animals, the FUS-induced BBB area
clearly covered the tumor tissue (Figure ;small(-day)or
large (-day) tumors). Sonication resulted in increased Gd-
DTPA accumulation in the tumor and in the peripheral BBB-
opened area as evidenced by signal enhancement in the R1-
map images. AUC maps showed maintained high staining
intensities aer Gd-DTPA injection. In the small (-day)
tumors, FUS resulted in an increase of R1signal from to
about . s−1 andanincreaseinGd-DTPAdepositionofabout
(from to ). However, in large (-day) tumors,
FUS did not lead to a signicant change in the R1-signal,
which increased from . to s−1,ortheAUCvaluewhich
increased by only about (from to ).
e kinetics of Gd-DTPA accumulation were evaluated
aer a single sonication treatment in thirty animals (normal
rats: 𝑛=18; small-tumor model: 𝑛=6;large-tumor
model: 𝑛=6). An ROI from the BBB-opened area on T2-
weighted images (target) and the corresponding ROI from
the contralateral brain (contralateral) were used to infer Gd-
DTPA concentration from the R1signals and the AUC over
time in the -day tumors (volume <. cm3), -day tumors
(volume >. cm3), or normal controls aer sonication
(Figure ). Figures (a)–(c) showed the comparison of
changes in R1as a function of time (from to min)
for three typical animals. When considering the peak value
over the whole scanning process, FUS caused the highest
enhancement in R1signal of contrast agent in normal tissue
(from . to .; Figure (a)). Sonication also led to a large
increase in R1signal in -day tumor, from . to . s−1
(Figure (b)). However, the already high permeability of -
day tumor to Gd-DTPA was not signicantly increased by
FUS, from . to . s−1 (Figure (c)).
e corresponding AUC (accumulation of R1)asa
function of time (from to min) in these three animals is
showninFigures(d)–(f). In the normal animal, total Gd-
DTPA accumulation in the BBB-opening area was increased
from . to . pmol by FUS (Figure (d)). Gd-DTPA
accumulation in the -day tumor increased from . to
. pmol, compared to only about pmol over time in
the contralateral control hemisphere (Figure (e)). However,
BioMed Research International
2.5
2
1.5
1
0.5
00 80 0 80 0 80
0.8
0.6
0.4
0.2
0
Time aer FUS (min)
Contralateral
FUS site
Normal
Gd-DTPA concentration (𝜇M)
R1(s−1)
Tumor, 10 d
Tumor, 17 d
(a)
Time aer FUS (min)
Normal
Ratio to control
8
6
4
2
0080
Tumor, 10 d
Tumor, 17 d
(b)
0
Contralateral
FUS site
Normal
5500
3500
1500
1200
800
400
0
Concentration by AUC (𝜇M)
R1AUC
Tumor, 10 d Tumor, 17 d
(c)
F : (a) Instantaneous R1measured immediately or minutes aer FUS exposure in normal animals, -day and -day tumor animals.
(b) Ratio of instantaneous R1values, immediately and min aer FUS. (c) Corresponding R1AUC of (a). Images were acquired in time slot
II.
Gd-DTPA accumulated to about the same high value with
and without FUS (. versus . pmol) in the -day
tumor (Figure (f )).
Gd-DTPA levels on the FUS-treated and contralateral
side were also evaluated at either min or min aer a
single sonication treatment (Figure ). e Gd-DTPA signal
intensity increased in the BBB-opening area min aer
sonication as evidenced by an increase in the R1signal from
. to . s−1 for control and from . to . s−1 for
small-tumor animals. However, the contrast agent signals
in the FUS-enhanced large tumors were the same as for
the contralateral region (about . s−1). All R1signals in
these brain tissues returned to baseline (about . s−1)at
min aer sonication. e ratio of R1between the FUS-
exposed and control areas went from . s−1 at min to
. s−1 atmininnormaltissueandfromabout.to
s
−1 for the tumor-bearing animals (Figure (b)). e AUC
at min and min was compared between control and
BBB-opened brain regions. At min aer sonication, the
accumulated Gd-DTPA concentration of the BBB-opened
area increased to . 𝜇M, compared to . 𝜇Mforthe
smalltumorandalimitedincreaseto𝜇Mforthelarge
tumor (Figure (c)).
Next we evaluated the correlation between EB leakage
(Figure (a)) and Gd-DTPA accumulation (estimated by
the AUC) in the same region of the brain. We found that
the accumulated distribution of Gd-DTPA (i.e., AUC) was
highly correlated with the distribution of EB (𝑟2= 0.8897)
(Figure (b)). us, R1-based pharmacodynamic analysis
providedareasonablemapofthepermeabilityoftheBBB-
disrupted region to EB dye over time. EB dye and Gd-
DTPA accumulation showed the same tendency of higher
overall permeability in tissues of large tumors and less depen-
dence on FUS treatment, as evidenced from the ratios of
accumulation between contralateral and FUS-treated regions
(Figure (c)).
Next, we analyzed Gd-DTPA deposition dynamics in
experimental group animals by AUC analysis for three
individual time slots: (I) before FUS, (II) immediately aer
FUS, and (III) hours aer FUS. Gd-DTPA concentration
BioMed Research International
Normal BBB group
Normal tumor group
Contralateral
FUS site
Tumo r s ite
Evans blue (𝜇M)
5
4
3
2
1
0
−10 200 400 600 800 1000
Gd-DTPA (𝜇M)
𝑦 = 0.0051𝑥 − 0.3963
𝑟2=0.8897
(a)
Evans blue (𝜇M)
12
10
8
6
4
2
0
Normal, contra
Normal, FUS
Normal, FUS
Tumor, 10 d
Tumor, 17 d
Normal
Tumor, 10 d
Tumor, 17 d
(b)
Normal, FUS
8
6
4
2
0
−102468
Ratio of Gd-DTPA
Ratio of Evans blue
Tumor, 10 d
Tumor, 17 d
(c)
F : (a) Correlation between quantied Evans blue concentration and R1-estimated Gd-DTPA concentration in normal (red circle) and
tumor (blue triangle) brains. Le arrow: contralateral side and right arrow: with FUS (red circle); normal tumor (blue triangle). (b) Evans
blue aer FUS exposure in the normal contralateral region, normal FUS targeting region, -day tumor region aer FUS, and -day tumor
region aer FUS. (c) Correlation of the ratios of FUS: contralateral concentrations of Evans blue and Gd-DTPA in normal and tumor-bearing
brains.
BioMed Research International
2.5
2
1.5
1
0.5
0
0.8
0.6
0.4
0.2
0
I/ before
FUS
II/ immediately
aer FUS
Normal
Gd-DTPA (𝜇M)
R1(s−1)
Tumor, 10 d
Tumor, 17 d
III/ 2 hours
aer FUS
(a)
10
7
4
1
Ratio
I/ before
FUS
II/ immediately
aer FUS
Normal
Small tumor
Large tumor
III/ 2 hours
aer FUS
(b)
5500
3500
1500
0I/ before
FUS
II/ immediately
aer FUS
Normal
Gd-DTPA accumulation (𝜇M)
1200
800
400
0
R1AUC
Tumor, 10 d
Tumor, 17 d
III/ 2 hours
aer FUS
(c)
10
7
4
1
I/ before
FUS
II/ immediately
aer FUS
Ratio
Normal
Small tumor
Large tumor
III/ 2 hours
aer FUS
(d)
F : (a) Instantaneous R1values of the normal animals, -day and -day tumor animals at MRI acquisition time slots I, II, and III. (b)
Ratio of instantaneous R1to the value in time slot I. (c) Corresponding R1AUCof(a).(d)RatioofR
1AUCtothevalueintimeslotI.
and accumulation in the target area presented the same trend
at all three time points (Figure ). As before, we observed a
transient peak of R1andtheAUCintheBBB-openedbrain
just aer sonication (Figures (a) and (c)). However, the
increase in the ratio of R1or AUC again diered between
the normal and tumor-bearing animals (Figures (b) and
(d)). e ratio was highest in the normal BBB-opening
area (R1signal . times than before FUS; AUC: . times),
followed by the small-tumor model (R1signal: . times;
AUC: . times), with no signicant change in the large
tumor, conrming our previous observations that FUS did
not signicantly further aect permeability in large tumors.
ese ratios subsequently decreased at the time point hours
aer FUS induction, returning to approximately the same
values of DCE-MRI as originally observed before sonication.
is result implied that at hours, the BBB-opening area had
recovered to the same baseline permeability level to contrast
agent as prior to sonication.
HE staining of tumors days aer implantation showed
even staining without scattered red blood cells in the absence
of FUS (Figures (a) and (b)). Tumor cells were charac-
terizedbydensenucleardistribution,andonlytinyareas
of gliosis inltrated with chronic inammatory cells and
some hemorrhaging were found (Figures (c) and (d)).
HE staining of tumors days aer implantation revealed
a number of regions with extensive apoptosis and cavities
inthecoreofthetumor,andhemorrhagicstructureswith
scattered and spreading erythrocytes could be observed
BioMed Research International
(a) (b) (c) (d)
(e) (f) (g)
100 𝜇m
(h)
F : HE staining. (a) and (b) Small (-day) tumor tissue without FUS exposure, x and x. (c) and (d) Small (-day) tumor tissue
with FUS exposure, x and x. (e) and (f) Large (-day) tumor tissue without FUS exposure, x and x. (c) and (d) Large (-day)
tumor tissue with FUS exposure, x and x.
5500
3500
1500
0
1200
800
400
0
Before
FUS
Immediately
aer FUS
Gd-DTPA concentration (𝜇M)
R1AUC
10-day tumor
17-day tumor
17-day tumor, peripheral
17-day tumor, core
2 hours
aer FUS
(a)
Immediately
aer FUS
R1concentration increase (%)
100
80
60
40
20
0
10-day tumor
17-day tumor
17-day tumor, peripheral
17-day tumor, core
2 hours
aer FUS
(b)
F : (a) R1AUC analysis of the core and peripheral subregions of -day tumor. (b) R1AUCincreaserelativetotimeslotI.
around discontinuous vasculature (Figures (e) and (f)),
supporting our ndings of high permeability of -day tumors
based on observation of Gd-DTPA deposition. is structure
did not change signicantly aer sonication (Figures (g) and
(h)) with hemorrhagic regions remaining similar to those in
the unexposed tumor.
Previous reports showed that the tumor core consists
of a bulky necrotic mass without functional vasculature,
whereas the tumor periphery maintains a high degree of
vasculature structure [,]. We therefore hypothesized that
microbubble-enhanced FUS exposure would have a bigger
eect on enhancing permeability in the peripheral tumor.
We further divided the -day tumor animals into core
and peripheral subregions and then repeated the MRI AUC
analysis (Figure ). We observed that, aer FUS exposure,
the -day peripheral tumor showed a similar trend to the -
day tumor, which showed a nearly .-fold of instantaneous
increase and % accumulation increase in Gd-DTPA, and
the permeability dropped signicantly two hours aer FUS
exposure to about half (%), which is similar to -day
tumor. In contrast, the -day tumor core mimicked the
behavior of the undivided -day tumor, accumulating high
BioMed Research International
levels of Gd-DTPA both before and aer FUS exposure
(AUC increase of only % aer FUS exposure) (Figure ).
is relatively low increase in the AUC persisted when the
tumor core was re-evaluated at hours aer FUS exposure
(%). ese observations suggest that the FUS-BBB opening
can provide the most pronounced drug delivery enhancing
eect on tumor peripheral or tumors with high-vascularity
stagetumors,yetonlyprovideslimitedeectsonbulkyand
necrotic tumors.
4. Discussion
is study demonstrated the pharmacodynamic characteris-
ticsofsmall-moleculeleakageatvariousstagesoftumorsaer
application of microbubble-enhanced FUS to open the BBB.
We analyzed two small molecules with similar molecular
weights to obtain complimentary data on pharmacodynamic
behavior. Gd-DTPA was used to provide contrast in MRI and
for semiquantitative verication of biodistribution in vivo,
andEBdyewasusedasameasureofdrugaccumulation
aer animal sacrice. ese two molecules, which normally
do not enter the brain parenchyma from the bloodstream,
could potentially be used as surrogate markers for drug
delivery. Although the dynamic distribution of Gd-DTPA
may dier from that of Evans blue, we demonstrated that the
AUC accumulation of Gd-DTPA analyzed by MRI was highly
correlated with EB accumulation in the brain (𝑟2= 0.8897),
implying that MRI AUC analysis of Gd-DTPA could predict
the concentration of EB accumulating in the brain, and may
thus have the potential to predict the pharmacodynamic
behavior and biodistribution of other therapeutic agents.
is study employed high-temporal-resolution dynamic
CE-MRI that could be utilized for small-molecule in vivo
distribution and semiquantication, as attempted in previous
studies [,]. e unique features provided by dynamic
CE-MRI include the capability of rapid evaluation and high
spatial resolution, as well as kinetic analysis to evaluate tumor
perfusion []. Positron emission tomography (PET) has
also been used for pharmacological studies in several tumor
types [,]. However, potential limitations of PET may
include its limited imaging resolution and complexity of
radiotracer synthesis. In small tumors, partial volume eects
may be signicant if the tumor size is less than twice the
resolution of the scanner []. MRI methods provide the
advantageofhavinggoodspatialresolutionequaltothat
of corresponding morphologic images. In addition, MRI
isminimallyinvasiveandposeslittlerisktopatients.We
used voxels of about . ×. ×mm
3to construct
images sucient for small-animal analysis. On the other
hand, PET relies on radiolabeled molecules that bind to
receptors to allow absolute quantication by detection of
isotopes. PET may also be limited by its high cost, limited
availability of radiotracer, and the need for a cyclotron as
well as onsite radiochemistry for radioisotope production
[]. PET involves comprehensive conjugation of radiotracers
and specic tailor-made molecules limiting its general use
for pharmacodynamic analysis. Although Gd-DTPA can-
not be directly conjugated to therapeutic molecules, the
detection of coadministered Gd-DTPA by CE-MRI is highly
correlated with targeted molecules, providing an excellent
tool for monitoring vasculature and evaluating tissue/tumor
permeability at high temporal/spatial resolution, suggesting
its continued usefulness for pharmacodynamic analysis of
brain drug delivery.
Tumor tissues are known to have high permeability due
to the presence of large endothelial cell gaps, incomplete
basement membrane, and the relative lack of pericyte or
smooth muscle association with endothelial cells [,].
In addition, the network of vasculature in solid tumors is
markedly dierent from the normal hierarchical branching
patterns and contains leaky vessel structures. Variations in
permeabilityarealsoassociatedwiththetumorgradeaswell
as various neoplastic eects that could disrupt the BBB [].
In this study, we observed that tumors with dierent levels of
progression showed dierent characteristics of blood-vessel
permeability and small-molecule accumulation. We found
that the AUC80 min in small tumors was 452 ± 122.5𝜇M,
whereas in large tumors it reached 754 ± 48.3𝜇M. More-
over, we conrmed that FUS-BBB opening provided a %
enhancement of accumulation of Gd-DTPA in small brain
tumors and a % enhancement at the large tumor periphery,
implying that FUS-BBB opening is an eective approach
to increase brain-tumor permeability and therefore enhance
delivery of therapeutic molecules.
Histological examination by HE staining showed that
smaller (-day) tumors had well-ordered vasculature with
fewer abnormal endothelial cell gaps (Figures (a) and (b)).
e blood vessel density in small tumor was lower, resulting
in less Gd-DTPA and EB accumulation. In contrast, -
day tumor tissues contained more large fenestrae (Figures
(e) and (f)), consistent with previous pathological ndings
that high-grade brain tumors contain neovasculature and
apoptotic tumor cells, leading to hyperpermeability [].
ese pathological changes are consistent with the increased
Gd-DTPA and EB accumulation that we observed in -day
tumor tissues before sonication.
Although AUC80 min correlated well with the pharmaco-
dynamic behavior of another small molecule (EB), Gd-DTPA
accumulation can be very dierent even under the same
FUS exposure conditions, for example, varying from .
to 𝜇M(Figure(b)).eselargevariationswerelikely
due to dierences in skull thickness and angle of incidence
between the FUS beam and the skull surface among the
animals [], and the presence of standing waves produced
in the skull cavity that alter the peak pressure at the target
position and thus the level of BBB opening []. Since FUS-
BBB opening may vary substantially, it is essential to perform
an AUC analysis during CE-MRI to monitor small-molecule
delivery into the brain for individual subjects and targets.
5. Conclusion
In this study, we characterized the dynamics of BBB opening
in normal and tumor tissues using DCE-MRI with Gd-
DTPA contrast agent, and related them to the concentrations
of Evans Blue determined from tissues aer sacrice. e
BioMed Research International
concentrations of the surrogate tracer (Gd-DTPA) and EB
dye showed a strong linear correlation. With this dynamic
information of tumor permeability, the pharmacodynamic
model can be modied to eventually take into account
parameters that aect drug delivery over time. Tumor periph-
eral or high-vascular tumor may have the most signicant
benet on blood-brain or blood-tumor permeability increase,
which gives critical information when intending to apply
FUS for brain drug enhanced delivery. We hope to use such
pharmacodynamic predictions along with FUS-induced BBB
opening to develop a method for image-guided drug delivery
that can estimate the amount of drug that will be delivered to
tissues at each time point.
Conflict of Interests
e authors declare that there is no conict of interests.
Acknowledgments
e authors thank the Molecular Imaging Center, Chang
Gung Memorial Hospital. is work was supported by the
National Science Council, Taiwan (Grant nos. --
B--, --M-A-, and --E---
MY).
References
[] K. Hynynen, N. McDannold, N. Vykhodtseva, and F. A. Jolesz,
“Noninvasive MR imaging-guided focal opening of the blood-
brain barrier in rabbits,” Radiology,vol.,no.,pp.–,
.
[]K.Hynynen,N.McDannold,N.A.Sheikov,F.A.Jolesz,and
N. Vykhodtseva, “Local and reversible blood-brain barrier
disruption by noninvasive focused ultrasound at frequencies
suitable for trans-skull sonications,” NeuroImage,vol.,no.,
pp. –, .
[] K.Hynynen,N.McDannold,N.Vykhodtsevaetal.,“Focaldis-
ruption of the blood-brain barrier due to -kHz ultrasound
bursts: a method for molecular imaging and targeted drug
delivery,” Journal of Neurosurgery,vol.,no.,pp.–,
.
[]M.Kinoshita,N.McDannold,F.A.Jolesz,andK.Hyny-
nen, “Targeted delivery of antibodies through the blood-brain
barrier by MRI-guided focused ultrasound,” Biochemical and
Biophysical Research Communications,vol.,no.,pp.–
, .
[] N. McDannold, N. Vykhodtseva, and K. Hynynen, “Eects of
acoustic parameters and ultrasound contrast agent dose on
focused-ultrasound induced blood-brain barrier disruption,”
Ultrasound in Medicine and Biology,vol.,no.,pp.–,
.
[]H.L.Liu,Y.Y.Wai,W.S.Chenetal.,“Hemorrhagedetec-
tion during focused-ultrasound induced blood-brain-barrier
opening by using susceptibility-weighted magnetic resonance
imaging,” Ultrasound in Medicine and Biology,vol.,no.,pp.
–, .
[]H.L.Liu,P.H.Hsu,P.N.Chuetal.,“Magneticresonance
imaging enhanced by superparamagnetic iron oxide particles:
usefulness for distinguishing between focused ultrasound-
induced blood-brain barrier disruption and brain hemorrhage,”
Journal of Magnetic Resonance Imaging,vol.,no.,pp.–,
.
[]M.Kinoshita,N.McDannold,F.A.Jolesz,andK.Hynynen,
“Noninvasive localized delivery of Herceptin to the mouse
brain by MRI-guided focused ultrasound-induced blood-brain
barrier disruption,” Proceedings of the National Academy of
Sciences of the United States of America,vol.,no.,pp.–
, .
[]L.H.Treat,N.McDannold,N.Vykhodtseva,Y.Zhang,K.
Tam, and K. Hynynen, “Targeted delivery of doxorubicin to
the rat brain at therapeutic levels using MRI-guided focused
ultrasound,” International Journal of Cancer,vol.,no.,pp.
–, .
[] J. Mei, Y. Cheng, Y. Song et al., “Experimental study on
targeted methotrexate delivery to the rabbit brain via mag-
netic resonance imaging-guided focused ultrasound,” Journal of
Ultrasound in Medicine, vol. , no. , pp. –, .
[] H. L. Liu, M. Y. Hua, P. Y. Chen et al., “Blood-brain bar-
rier disruption with focused ultrasound enhances delivery of
chemotherapeutic drugs for glioblastoma treatment,” Radiol-
ogy,vol.,no.,pp.–,.
[] W. M. Pardridge, “e blood-brain barrier: bottleneck in brain
drug development,” NeuroRx,vol.,no.,pp.–,.
[] A. Jemal, R. Siegel, E. Ward et al., “Cancer statistics, ,” CA:
A Cancer Journal for Clinicians,vol.,no.,pp.–,.
[] S. A. Grossman and J. F. Batara, “Current management of
glioblastoma multiforme,” Seminars in Oncology,vol.,no.,
pp.–,.
[]L.H.Treat,N.McDannold,Y.Zhang,N.Vykhodtseva,
and K. Hynynen, “Improved anti-tumor eect of liposomal
doxorubicin aer targeted blood-brain barrier disruption by
MRI-guided focused ultrasound in rat glioma,” Ultrasound in
Medicine and Biology,vol.,no.,pp.–,.
[] H. L. Liu, M. Y. Hua, H. W. Yang et al., “Magnetic resonance
monitoring of focused ultrasound/magnetic nanoparticle tar-
geting deliver y of therapeutic agents to the brain,” Proceedings of
the National Academy of Sciences of the United States of America,
vol.,no.,pp.–,.
[] K. Liu, G. Wen, X. F. Lv et al., “MR imaging of cerebral extraven-
tricular neurocytoma: a report of cases,” e American Journal
of Neuroradiology,.
[] N. McDannold, N. Vykhodtseva, and K. Hynynen, “Blood-
brain barrier disruption induced by focused ultrasound and cir-
culating preformed microbubbles appears to be characterized
by the mechanical index,” Ultrasound in Medicine and Biology,
vol.,no.,pp.–,.
[] N. Vykhodtseva, N. McDannold, and K. Hynynen, “Progress
and problems in the application of focused ultrasound for
blood-brain barrier disruption,” Ultrasonics,vol.,no.,pp.
–, .
[]F.Y.Yang,W.M.Fu,W.S.Chen,W.L.Yeh,andW.L.Lin,
“Quantitative evaluation of the use of microbubbles with tran-
scranial focused ultrasound on blood-brain-barrier disruption,”
Ultrasonics Sonochemistry,vol.,no.,pp.–,.
[] F.Y.Yang,W.M.Fu,R.S.Yang,H.C.Liou,K.H.Kang,and
W. L. Lin, “Quantitative evaluation of focused ultrasound with
a contrast agent on blood-brain barrier disruption,” Ultrasound
in Medicine and Biology,vol.,no.,pp.–,.
[] J. J. Choi, J. A. Feshitan, B. Baseri et al., “Microbubble-size
dependence of focused ultrasound-induced bloodBrain barrier
BioMed Research International
opening in mice in vivo,” IEEE Transactions on Biomedical
Engineering,vol.,no.,pp.–,.
[] R. Chopra, N. Vykhodtseva, and K. Hynynen, “Inuence of
exposure time and pressure amplitude on blood-brain-barrier
opening using transcranial ultrasound exposures,” ACS Chemi-
cal Neuroscience,vol.,no.,pp.–,.
[] J.Park,Y.Zhang,N.Vykhodtseva,F.A.Jolesz,andN.J.McDan-
nold, “e kinetics of blood brain barrier permeability and
targeted doxorubicin delivery into brain induced by focused
ultrasound,” Journal of Controlled Release,vol.,no.,pp.–
, .
[] P. S. Tos and A. G. Kermode, “Measurement of the blood-
brain barrier permeability and leakage space using dynamic
MR imaging. . Fundamental concepts,” Magnetic Resonance in
Medicine,vol.,no.,pp.–,.
[] F. Vlachos, Y. S. Tung, and E. Konofagou, “Permeability
dependence study of the focused ultrasound-induced blood-
brain barrier opening at distinct pressures and microbubble
diameters using DCE-MRI,” Magnetic Resonance in Medicine,
vol. , no. , pp. –, .
[] I. M. Noebauer-Huhmann, P. Szomolanyi, V. Juras, O. Kra,
M. E. Ladd, and S. Trattnig, “Gadolinium-based magnetic
resonance contrast agents at tesla: in vitro T relaxivities in
human blood plasma,” Investigative Radiology,vol.,no.,pp.
–, .
[] E. A. Neuwelt, P. A. Barnett, C. I. McCormick, L. G. Remsen, R.
A. Kroll, and G. Sexton, “Dierential permeability of a human
brain tumor xenogra in the nude rat: impact of tumor size and
method of administration on optimizing delivery of biologically
diverse agents,” Clinical Cancer Research,vol.,no.,pp.–
, .
[] H. Sarin, A. S. Kanevsky, H. Wu et al., “Eective transvascular
delivery of nanoparticles across the blood-brain tumor barrier
into malignant glioma cells,” JournalofTranslationalMedicine,
vol. , article , .
[] A. R. Padhani and J. E. Husband, “Dynamic contrast-enhanced
MRI studies in oncology with an emphasis on quantication,
validation and human studies,” Clinical Radiology,vol.,no.
, pp. –, .
[] P. S. Tos, G. Brix, D. L. Buckley et al., “Estimating kinetic
parameters from dynamic contrast-enhanced T()-weighted
MRI of a diusable tracer: standardized quantities and sym-
bols,” Journal of Magnetic Resonance Imaging,vol.,no.,pp.
–, .
[] H. Olsson, C. Halldin, C. G. Swahn, and L. Farde, “Quan-
tication of [C]FLB binding to extrastriatal dopamine
receptors in the human brain,” JournalofCerebralBloodFlow
and Metabolism,vol.,no.,pp.–,.
[] K.W.Kang,D.S.Lee,J.H.Choetal.,“QuanticationofF-
FDG PET images in temporal lobe epilepsy patients using
probabilistic brain atlas,” NeuroImage,vol.,no.,part,pp.
–, .
[] S. L. Bacharach, S. K. Libutti, and J. A. Carrasquillo, “Measur-
ing tumor blood ow with H215O: practical considerations,”
Nuclear Medicine and Biology, vol. , no. , pp. –, .
[] M. S. Alavijeh and A. M. Palmer, “Measurement of the pharma-
cokinetics and pharmacodynamics of neuroactive compounds,”
Neurobiology of Disease,vol.,no.,pp.–,.
[]H.F.Dvorak,L.F.Brown,M.Detmar,andA.M.Dvorak,
“Vascular permeability factor/vascular endothelial growth fac-
tor, microvascular hyperpermeability, and angiogenesis,” e
American Journal of Pathology,vol.,no.,pp.–,
.
[] L. E. Benjamin, D. Golijanin, A. Itin, D. Pode, and E. Keshet,
“Selective ablation of immature blood vessels in established
humantumorsfollowsvascularendothelialgrowthfactorwith-
drawal,” Journal of Clinical Investigation,vol.,no.,pp.–
, .
[] J. S. Taylor, P. S. Tos, R. Port et al., “MR imaging of tumor
microcirculation: promise for the new millennium,” Journal of
Magnetic Resonance Imaging, vol. , no. , pp. –, .
[] L. F. Brown, K. T. Yeo, B. Berse et al., “Expression of vascular
permeability factor (vascular endothelial growth factor) by
epidermal keratinocytes during wound healing,” Journal of
Experimental Medicine,vol.,no.,pp.–,.
[] M. A. O’Reilly, A. Muller, and K. Hynynen, “Ultrasound
insertion loss of rat parietal bone appears to be proportional
to animal mass at submegahertz frequencies,” Ultrasound in
Medicine and Biology, vol. , no. , pp. –, .
[] M. A. O’Reilly, Y. Huang, and K. Hynynen, “e impact
of standing wave eects on transcranial focused ultrasound
disruption of the blood-brain barrier in a rat model,” Physics
in Medicine and Biology,vol.,no.,pp.–,.
Available via license: CC BY
Content may be subject to copyright.
Content uploaded by Po-chun Chu
Author content
All content in this area was uploaded by Po-chun Chu
Content may be subject to copyright.
