Download Multiparametric Monitoring of Tumor Response to

Published Online First on January 13, 2009 as 10.1158/0008-5472.CAN-08-2001
Research Article
Multiparametric Monitoring of Tumor Response to
Chemotherapy by Noninvasive Imaging
Zdravka Medarova, Leonid Rashkovetsky, Pamela Pantazopoulos, and Anna Moore
Molecular Imaging Laboratory, MGH/HST Athinoula A. Martinos Center for Biomedical Imaging, Department
of Radiology, Massachusetts General Hospital/Harvard Medical School, Boston, Massachusetts
underglycosylated on more than 90% of breast tumors and whose
abundance seems to be linked to tumor progression and response
to chemotherapy (1). Our tumor-specific multimodal imaging
probe consists of superparamagnetic iron oxide nanoparticles
(MN) for magnetic resonance imaging (MRI) modified with Cy5.5
dye [for optical near-IR fluorescence (NIRF) imaging] and has
peptides (EPPT) specifically recognizing uMUC-1 attached to its
dextran coat (1). The dual-modality approach combines the highspatial resolution, tomographic capability, and unlimited tissue
penetration of MRI with the high sensitivity and low cost of in vivo
optical imaging as a validation tool (2).
In the past, we have done extensive studies validating the utility
of MUC-1 as a candidate antigen and the specificity of MN-EPPT as
a multimodal targeted probe in various cancer models (1, 3). In
our new studies described here, we used this imaging probe for
the simultaneous quantitative assessment of tumor volume and
target antigen availability in breast tumors before and after
chemotherapy. The importance of these studies is underscored
by the fact that breast cancer is the second leading cause of cancerrelated deaths in women of the western world. In the United States
alone, more than 180,000 new cases are diagnosed each year. Of
these patients, f25% will succumb to their disease despite early
diagnosis and aggressive therapeutic intervention, possibly due to
the inability to optimally tailor-fit treatments on an individualized
basis.
As of today, the most reliable predictor of overall breast cancer
survival and therapeutic outcome in breast cancer patients is the
stage at which the tumor is diagnosed. At the same time, early
therapeutic intervention holds promise because of its life-saving
potential. Therefore, it is critical to devise ways in which to explore
the molecular complexity of breast cancer for early diagnosis
and monitoring of treatment progress. There is a wealth of
imaging modalities capable of evaluating the tumor and the
tumoral response to therapy, including mammography and
sonography (4–7), MRI (8–11), and scintigraphy or positron
emission tomography (12–14). These modalities, however, primarily evaluate tumor volume and metabolism but do not provide
dynamic information about the molecular profile of the lesion,
which would be a much earlier and more reliable predictor of
tumor outcome (4).
Here we have addressed the need to develop new tumor
antigen–specific contrast agents in combination with highresolution in vivo imaging methods for the monitoring of response
to therapy. We believe that this study is significant because
it explores a noninvasive imaging approach to monitor change
in tumor size and the relative availability of a tumor antigen
(uMUC-1). uMUC-1 provides a unique advantage in this case
because it shows predictable changes in phenotype directly linked
to breast cancer progression as well as to therapeutic outcome (15).
Considering that related iron oxides are already in clinical use (16),
if successful, our studies would establish the feasibility of a
Abstract
With the emerging concept of individualized cancer therapy, it
becomes crucial to develop methods for the noninvasive
assessment of treatment outcome. With this in mind, we
designed a novel approach for the comprehensive evaluation
of response to chemotherapy with the established agent
doxorubicin in a preclinical breast cancer model. This
approach delivers information not only about change in
tumor size but also about target antigen expression. Our
strategy relies on a tumor-specific contrast agent (MN-EPPT)
targeting the underglycosylated MUC-1 (uMUC-1) tumor
antigen, found on more than 90% of breast cancers and
predictive of chemotherapeutic response. MN-EPPT consists of
superparamagnetic iron oxide nanoparticles (MN) for magnetic resonance imaging, modified with Cy5.5 dye ( for near-IR
fluorescence optical imaging), and conjugated to peptides
(EPPT), specifically recognizing uMUC-1. In vivo, treatment of
mice bearing orthotopic human breast carcinomas with
doxorubicin led to a reduction in tumor mass and resulted
in down-regulation of uMUC-1. The tumor-specific accumulation of MN-EPPT allowed the assessment of change in tumor
volume by noninvasive imaging. Furthermore, in mice injected
with MN-EPPT, tumor delta-T2 was significantly reduced after
treatment with doxorubicin, indicating a lower accumulation
of MN-EPPT and reflecting the reduced expression of uMUC-1.
With these studies, we have shown the utility of magnetic
resonance imaging for the multiparametric characterization
of breast tumor response to chemotherapy. This approach has
the potential of significantly advancing our ability to better
direct the development of molecularly targeted individualized
therapy protocols because it permits the monitoring of
therapy on a molecular scale. [Cancer Res 2009;69(3):OF1–8]
Introduction
The emergence of molecularly targeted cancer therapies
mandates the development of methods to directly determine their
efficacy. Noninvasive imaging techniques are currently available
for visualizing different pathologic conditions. However, their use
for cancer monitoring is limited due to the lack of tumor-specific
imaging probes. We have previously developed a multimodal
imaging probe (MN-EPPT) targeting the underglycosylated
mucin-1 tumor antigen (uMUC-1), which is overexpressed and
Requests for reprints: Anna Moore, MGH/MIT/HMS Athinoula A. Martinos Center
for Biomedical Imaging, Department of Radiology, Massachusetts General Hospital/
Harvard Medical School, Room 2301, Building 149, 13th Street, Charlestown, MA 02129,
Phone: 617-724-0540; Fax: 617-726-7422; E-mail: [email protected].
I2009 American Association for Cancer Research.
doi:10.1158/0008-5472.CAN-08-2001
www.aacrjournals.org
OF1
Cancer Res 2009; 69: (3). February 1, 2009
Cancer Research
scanner equipped with ParaVision 3.0 software. The imaging protocol
consisted of coronal T2-weighted spin echo (SE) pulse sequences with the
following parameters: SE TR/TE = 3,000/[8, 16, 24, 32, 40, 48, 56, 64]; field of
view, 40 40 mm; matrix size, 128 128 pixels; slice thickness, 0.5 mm; inplane resolution, 312 312 Am. Image reconstruction and analysis were
done using Marevisi 3.5 software (Institute for Biodiagnostics, National
Research Council, Canada). T2 maps were constructed according to
established protocol by fitting the T2 values for each of the eight echo
times (TE) to a standard exponential decay curve. T2 relaxation times were
calculated by manually segmenting out the cell pellet on magnetic
resonance images.
Flow cytometry. Flow cytometry was done to confirm if the MN-EPPT
specificity for uMUC-1 translates in reduced probe accumulation in breast
adenocarcinoma cells, following uMUC-1 down-regulation by doxorubicin.
Flow cytometry was done on cells treated as described in ‘‘In vitro cell
treatment,’’ using a FACSCalibur (Becton Dickinson) equipped with the
Cell Quest software package (Becton Dickinson).
Tumor model. To establish a preclinical orthotopic tumor model of
human breast cancer, 5- to 6-wk-old female nu/nu mice (n = 10;
Massachusetts General Hospital Radiation Oncology breeding facilities)
were inoculated in the right mammary fat pad with the uMUC-1–positive
human breast adenocarcinoma cell line BT-20 (American Type Culture
Collection), as previously described (18).
All animal experiments were done in compliance with institutional
guidelines and according to the animal protocol approved by the
Subcommittee on Research Animal Care at Massachusetts General Hospital.
Doxorubicin treatment. To establish a clinically relevant treatment
model, we used the standard chemotherapeutic agent doxorubicin. The
treatment protocol involved i.v. injections of 7 mg/kg doxorubicin in saline
solution once weekly for 2 consecutive weeks beginning f8 d after tumor
implantation once tumors had reached a diameter of 0.5 cm, as previously
suggested (17). Saline solution–injected animals served as nontreated
controls. As previously described, to evaluate tumor response to
chemotherapy, in vivo imaging was done 1 d before the beginning of
treatment and 1 d after the completion of treatment (3).
In vivo MRI. MRI was done before and 24 h after i.v. injection of MNEPPT or the scrambled control probe, MN-SCR (10 mg Fe/kg), using a 9.4T
Bruker horizontal bore scanner equipped with ParaVision 3.0 software using
sequences as for in vitro MRI. Image reconstruction and analysis were done
as described for in vitro MRI.
Tumor volumes and tumor T2 relaxation times were calculated by
manually segmenting out the tumor on magnetic resonance images.
Quantitative evaluation of differential tumor growth by MRI was based on
multislice T2-weighted images. The volume was estimated according to the
formula for the volume of an ellipsoid: V = 4/3p(abc), where a and b are the
equatorial radii (along the x and y axes) and c is the polar radius (along the
z axis). For T2 map analysis of relaxation times, the terminal slices were not
included in the analysis to avoid interference from partial volume effects.
Relative MN-EPPT accumulation in the tumors was estimated based on the
following formula: T2 before injection
T2 after injection (delta-T2, in
milliseconds).
In vivo and ex vivo optical imaging. In vivo NIRF optical imaging was
done immediately after each MRI session. Animals were placed into a
whole-mouse imaging system (Imaging Station IS2000MM, Eastman Kodak
Company) and imaged in the Cy5.5 channel. At the end point of each
experiment following the last imaging session, mice were sacrificed; tumors
were excised, placed in the optical imaging system, and imaged ex vivo.
Image analysis was done using the Kodak 1D 3.6.3 Network software. The
actual volumes of excised tumors were determined by measuring tumor
dimensions ex vivo using calipers.
Immunohistochemistry and in situ apoptosis detection. To detect
the accumulation of MN-EPPT in tumors at the microscopic level, we
performed correlative immunohistochemistry. Tumors were embedded in
Tissue-Tek optimum cutting temperature compound (Sakura Fineteck) and
snap-frozen in liquid nitrogen. Tumors were then cut into 7-Am frozen
sections, fixed in 2% paraformaldehyde, washed, counterstained with
Vectashield mounting medium with 4¶,6-diamidino-2-phenylindole
multiparameter method for monitoring breast cancer response to
chemotherapy, based on the expression of molecular markers
implicated in disease progression.
Materials and Methods
Probe synthesis and characterization. The uMUC-1–targeted MNEPPT probe and the scrambled control probe MN-SCR were synthesized
and characterized as described (1, 3). Peptide sequences were as follows:
EPPT, C-AHA-A-R-E-P-P-T-R-T-F-A-Y-W-G-K(FITC); SCR, C-AHA-A-E-G-R-PT-F-P-T-R-A-Y-W-K(FITC). Synthesis resulted in a triple-labeled nanoparticle, consisting of FITC on EPPT peptide ( for fluorescence microscopy),
superparamagnetic iron oxide (MN, a magnetic label for MRI), and Cy5.5
dye (Amersham Biosciences) attached to the MN ( for NIRF optical
imaging). Iron concentration and peptide/FITC and Cy5.5 payloads were
determined as described in ref. 1. The resultant probe had an average iron
concentration of 7.3 F 0.16 mg/mL, 2.1 Cy5.5 molecules per nanoparticle,
and 4.43 peptides per nanoparticle.
In vitro cell treatment. To validate the observation that treatment with
doxorubicin down-regulates uMUC-1 in BT-20 breast adenocarcinoma cells,
cells were incubated with 0.4 Amol/L of doxorubicin-HCl (Sigma) for 48 h, as
previously described (17). PBS-treated cells served as controls. Following
incubation, the cells were analyzed for uMUC-1 expression by quantitative
reverse transcription-PCR (RT-PCR) analysis and Western blot, as described
below. Alternatively, to assess the relative accumulation of MN-EPPT in
these cells, following treatment with doxorubicin, the cells were incubated
at 37jC overnight with MN-EPPT or the scrambled control probe (MNSCR), washed, fixed in 2% paraformaldehyde, suspended in 0.5-mL PCR
tubes, and imaged by MRI, as described below. Following the MRI session,
the cells were transferred to fluorescence-activated cell sorting tubes and
analyzed by flow cytometry.
Real-time quantitative RT-PCR. Total RNA was extracted from breast
adenocarcinoma cells and tumor tissues using the RNeasy Mini Kit
according to the manufacturer’s protocol (Qiagen, Inc.). Relative levels of
uMUC-1 mRNA were determined by real-time quantitative RT-PCR
(TaqMan protocol). TaqMan analysis was done using an ABI Prism 7700
sequence detection system (PE Applied Biosystems). The PCR primers and
TaqMan probe specific for MUC-1 mRNA were designed using Primer
express software 1.5. Primer and probe sequences were as follows: forward
primer, 5¶-ACAGGTTCTGGTCATGCAAGC-3¶ (nucleotides 64–84 in the 5¶
nonrepetitive region); reverse primer, 5¶-CTCACAGCATTCTTCTCAGTAGAGCT-3¶ (nucleotides 139–164 in the 5¶ nonrepetitive region); and TaqMan
probe, 5¶-FAM-TGGAGAAAAGGAGACTTCGGCTACCCAGA-TAMRA-3¶
(nucleotides 96–124 in the 5¶ nonrepetitive region).
Eukaryotic 18S rRNA TaqMan PDAR Endogenous Control reagent mix
(PE Applied Biosystems) was used to amplify 18S rRNA as an internal
control according to the manufacturer’s protocol.
Western blot. For Western blot, total cell extracts were prepared by
solubilization of BT-20 breast adenocarcinoma cells with radioimmunoprecipitation assay lysis and extraction buffer (Pierce) with added Halt
Protease Inhibitor Cocktail (Pierce).
Protein concentration in the extracts was determined using the BCA
Protein Assay kit (Pierce). Samples were incubated with 1% SDS and 3% 2mercaptoethanol at 100jC for 5 min and centrifuged. Approximately 100 Ag
of total protein were separated on a 12% polyacrylamide gel (SDS-PAGE).
Precision Plus Protein Kaleidoscope Standards (Bio-Rad) were used as
molecular weight standards. Separated proteins were transferred onto a
polyvinylidene difluoride membrane and developed with One-Step Western
Advanced Kit for mouse primary antibody (GenScript Corp.). Primary anti–
Muc-1 mouse monoclonal antibodies specific for the backbone peptide
APDTRPAP (VU4H5 clone, Santa Cruz Biotechnology) and anti-tubulin
mouse monoclonal antibodies as an internal standard (V10178, Biomeda)
were used at 200- and 10,000-fold dilutions, respectively.
MRI of cell phantoms. To show that the specificity of MN-EPPT for
uMUC-1 can be used to probe for the down-regulation of uMUC-1 by
doxorubicin, we performed MRI of cell phantoms prepared as described in
the previous section. Imaging was done using a 9.4T Bruker horizontal bore
Cancer Res 2009; 69: (3). February 1, 2009
OF2
www.aacrjournals.org
Monitoring Tumor Response to Chemotherapy
Statistical analysis. All data were represented as mean F SE. Statistical
analysis was done using two-tailed Student’s t test and linear regression
where indicated. P V 0.05 was considered statistically significant.
Results
MN-EPPT accumulation in an orthotopic model of human
breast adenocarcinoma can be detected by MRI and NIRF
imaging. In this study, we used a preclinical orthotopic model of
human breast cancer to obtain information about tumor response
to the standard chemotherapeutic agent doxorubicin. As a first
step toward this goal, we evaluated whether MN-EPPT accumulates
in the orthotopic breast tumor model and whether this accumulation is sufficient to generate detectable contrast on T2-weighted
magnetic resonance and optical images. Eight days after tumor
implantation once tumors had reached a diameter of 0.5 cm,
female nu/nu mice were injected with the uMUC-1–targeted MNEPPT probe. Twenty-four hours after probe injection, mice were
subjected to MRI and fluorescence optical imaging. Representative
T2-weighted images and their corresponding T2 maps before and
24 hours after MN-EPPT injection are shown in Fig. 1. Before
injection of the contrast agents, breast tumors appeared with
characteristically long T2s (T2 = 46.8 F 2.0 ms). After administration of MN-EPPT, there was a significant decrease in tumor T2
relaxation time (T2 = 30.9 F 1.2 ms) compared with the precontrast value (P < 0.0011, n = 9; Fig. 1). In vivo NIRF optical
imaging was done immediately after the post-contrast MRI session.
In mice injected with MN-EPPT, there was a high-intensity
fluorescence signal coming from the implanted tumor (4,388.8 F
420.4 relative fluorescence units; Fig. 2A). Ex vivo imaging of
excised tumors (Fig. 2B) revealed bright fluorescence associated
with the tumor compared with muscle tissue, which was used to
define background fluorescence. Furthermore, ex vivo fluorescence
microscopy on frozen tumor sections revealed very extensive
accumulation of the probe in tumor cells (Fig. 2C). The good
Figure 1. MRI of MN-EPPT accumulation in orthotopic breast adenocarcinomas.
Representative T2 maps of animals bearing orthotopically implanted BT-20
human breast adenocarcinoma before (pre-contrast; left) and 24 h after
(post-contrast; right) i.v. injection of MN-EPPT. On pre-contrast T2-weighted
images and T2 maps, tumors (outlined) exhibit long T2s (high signal intensity).
On post-contrast images, there is a distinct shortening of the tumor T2 (low signal
intensity). Twenty-four hours after administration of MN-EPPT, there was a
significant (33.9 F 3.8%, P < 0.0011, n = 9) reduction in average T2 relaxation
times associated with the tumor.
(DAPI; Vector), and analyzed by fluorescence microscopy. Microscopy was
done using a Nikon Eclipse 50i fluorescence microscope equipped with an
appropriate filter set (Chroma Technology Corporation). Images were
acquired using a charge-coupled device camera with near-IR sensitivity
(SPOT 7.4 Slider RTKE, Diagnostic Instruments) and analyzed using SPOT
4.0 advanced version software (Diagnostic Instruments). Fluorescence was
collected in the green channel for detection of the FITC label on EPPT
peptides, in the blue channel for DAPI, and in the NIR channel for detection
of the Cy5.5 label on MN nanoparticles.
To evaluate levels of apoptosis in tumor cells, we performed a terminal
deoxynucleotidyl transferase–mediated deoxynucleotide triphosphate nick
end-labeling (TUNEL) assay (Apoptag Fluorescein In Situ Apoptosis
Detection kit, Chemicon International) according to the manufacturer’s
protocol. The nuclei were counterstained with DAPI and examined under
the fluorescence microscope.
Figure 2. Optical imaging of MN-EPPT
accumulation in orthotopic breast
adenocarcinomas. A, white-light image
(left), raw fluorescence image (middle ),
and a color-coded map (right ) of nu/nu
mice bearing orthotopically implanted
BT-20 human breast adenocarcinoma
obtained 24 h after i.v. injection of
MN-EPPT. Note the strong fluorescence
signal coming from the tumor on NIRF
images. B, ex vivo white-light image (left),
NIRF image (middle ), and color-coded
map (right ) of an excised breast tumor
24 h after injection of MN-EPPT. The
fluorescence originating from the tumor
was significantly higher than the signal
coming from excised muscle tissue, which
was used to define background. C,
fluorescence microscopy of frozen tumor
sections. Note the good colocalization
of FITC (EPPT) and Cy5.5 (MN)
fluorescence. Bar, 20 Am.
www.aacrjournals.org
OF3
Cancer Res 2009; 69: (3). February 1, 2009
Cancer Research
Figure 3. Tracking changes in tumor size following chemotherapy. A, representative color-coded T2 maps of nu/nu mice bearing orthotopic breast adenocarcinomas
before (top ) and after (bottom ) treatment with doxorubicin (DOX ). Tumor growth in mice treated with doxorubicin (left ) was compared with saline-treated controls
(right ). In treated animals, there was a reduction in tumor size by 53%. In saline-treated controls, there was a significant increase in tumor volume (P < 0.0011, n = 6).
B, representative white-light and NIRF images of nu/nu mice bearing orthotopic breast tumors before (top ) and after (bottom ) treatment with doxorubicin. Tumor
growth in doxorubicin-treated mice (left) was compared with that in saline-treated controls (right ). In treated animals, there was a 51% reduction in the ROI defined by
bright NIR fluorescence (NIRF ROI) and representative of tumoral probe accumulation. In saline-treated controls, there was a significant increase in the NIRF ROI
(P < 0.001, n = 6). C, linear regression analysis establishing correlation between tumor volume measurements by MRI and tumor ROI area measurements by
fluorescence optical imaging. Data points are derived from pretreatment and posttreatment tumor measurements and include both treated and nontreated control
animals (R 2 = 0.89, n = 18). D, ex vivo validation of doxorubicin treatment efficacy in vivo. Tumors were excised and photographed. Tumor sizes in mice treated with
doxorubicin were noticeably smaller than in saline-treated controls.
in saline-treated controls, tumor volume increased 9-fold by the
end of the experiment (P = 0.0011, n = 6; Fig. 3A).
These findings were independently confirmed by NIRF optical
imaging. In this case, the area characterized by near-IR fluorescence on the NIRF images served as a surrogate for the definition
of tumor margins. Visual inspection of NIRF images clearly
suggested a differential in tumor growth between experimental
and control animals (Fig. 3B). Semiquantitative analysis of relative
tumor size was done by selecting a ROI based on the section of
NIRF fluorescence and measuring the area of the ROI. Whereas in
animals treated with doxorubicin there was a decrease in relative
tumor size of 51%, in saline-treated controls tumor size increased
dramatically (P < 0.001, n = 6; Fig. 3B). In terms of relative
measurements of change in tumor size, the two modalities were
highly correlated (R 2 = 0.89; Fig. 3C) and accurately defined true
differences between experimental and control tumors, as seen
ex vivo (Fig. 3D).
Tracking change in tumor antigen expression following
chemotherapy in a preclinical orthotopic model of breast
cancer. There is evidence in the literature that treatment of breast
tumor cells with doxorubicin leads to down-regulation of the
uMUC-1 antigen (19). With this knowledge, we speculated that,
because the accumulation of MN-EPPT in tumors is predicated on
its specificity for the uMUC-1 tumor-specific antigen (1, 3) and the
change in T2 relaxation times post-contrast versus pre-contrast
(delta-T2) is a function of local contrast agent distribution and
abundance (20), by measuring delta-T2 relaxation times we would
colocalization of fluorescence in the NIR channel (Cy5.5, MN) and
in the green channel (FITC, EPPT peptide) indicated integrity of the
probe after persistence in the circulation (Fig. 2C).
With these experiments, we established that MN-EPPT accumulated in orthotopic breast tumor models and that this
accumulation can be detected by noninvasive MRI and NIRF
optical imaging. These findings set the stage for our next studies in
which we assessed the utility of the described imaging approach for
the monitoring of tumor response to chemotherapy. Because
the affinity of MN-EPPT for tumors is a function of its uMUC1–targeted nature (1, 3), we speculated that the uptake of this
probe by tumors would permit not only tumor delineation for the
calculation of tumor volume but also measurement of target
antigen expression as reflected by tumor T2 relaxation times.
Tracking change in tumor size following chemotherapy in a
preclinical orthotopic model of breast cancer. To determine if
MN-EPPT can serve as a tumor-targeted agent for the monitoring
of tumor response to chemotherapy, we treated tumor-bearing
animals with doxorubicin and compared changes in tumor size as
measured by MRI and NIRF optical imaging between experimental
mice and saline-treated controls. As seen in Fig. 3, whereas in
animals treated with doxorubicin there was an actual reduction in
tumor volume following the course of therapy, in saline-treated
controls tumor volume increased dramatically. These changes
were clearly detectable by MRI (Fig. 3A). Quantitative region of
interest (ROI) analysis of relative tumor volume revealed an
average decrease of 53% in doxorubicin-treated mice. By contrast,
Cancer Res 2009; 69: (3). February 1, 2009
OF4
www.aacrjournals.org
Monitoring Tumor Response to Chemotherapy
Next, we wanted to see if we could use the uMUC-1 selectivity of
MN-EPPT to probe for these changes. Populations of cells treated
with doxorubicin or PBS as described above were incubated with
PBS, MN-EPPT, or a scrambled control probe, MN-SCR (3), which
is identical to MN-EPPT except that the targeting EPPT peptide
is replaced with a scrambled peptide. Following incubation,
cell phantoms were imaged by MRI. As seen in Fig. 5A and B,
cells incubated with MN-EPPT displayed significantly lower T2
relaxation times than cells incubated with MN-SCR, further
confirming the specificity of MN-EPPT for uMUC-1 (P < 0.0001,
n = 6; Fig. 5A and B). Treatment with doxorubicin resulted in
significantly higher T2 relaxation times of MN-EPPT–incubated
cells (P = 0.0003, n = 6; Fig. 5A and B), indicating a lower
accumulation of the probe and closely mirroring the reduced
expression of the antigen in these cells as shown by quantitative
RT-PCR and Western blot. Unlike in cells incubated with MN-EPPT,
in cells incubated with the MN-SCR probe, treatment with
doxorubicin did not change the T2 relaxation times of the cells
(Fig. 5A and B).
The findings by MRI were also confirmed using flow cytometry
in the FL4 channel (Cy5.5), done immediately after the MRI session
(Fig. 5C and D). The relative fluorescence of cells incubated with
MN-EPPT was significantly higher than that of cells incubated with
MN-SCR (P = 0.007; Fig. 5C and D), reflective of the specificity of
the probe for uMUC-1. In addition, relative fluorescence was lower
in cells treated with doxorubicin and further incubated with MNEPPT than in control-treated cells (P = 0.007; Fig. 5C and D),
be able to indirectly extract information about the relative
availability of uMUC-1 as a result of chemotherapeutic treatment.
Obtaining such information would be very valuable from a clinical
point of view because changes in uMUC-1 expression directly relate
to tumor progression, response to therapy, metastatic potential,
and survival (21–24).
We first validated the observation that treatment with
doxorubicin down-regulated uMUC-1 in BT-20 breast adenocarcinoma cells. Cells were incubated with doxorubicin for 48 hours as
previously described (17). Following incubation, the cells were
analyzed for uMUC-1 expression by quantitative RT-PCR analysis
and Western blot. Both methods confirmed substantial downregulation of MUC-1 in BT-20 cells following treatment with
doxorubicin (Fig. 4A and B). Quantitative RT-PCR revealed a
significant 60 F 0.7% down-regulation of the gene (P = 0.028;
Fig. 4A). This down-regulation effect was accompanied by the
induction of considerable levels of apoptosis in the doxorubicintreated cells as assessed using TUNEL assay (Fig. 4C). Consistent
with the known mechanisms of cytotoxicity of doxorubicin, there
was abundant nuclear as well as extranuclear TUNEL staining,
indicating the formation of multiple micronuclei and loss of
nuclear membrane integrity. These events accompany mitotic
catastrophe and reflect advanced stages of doxorubicin-induced
apoptosis (25) and successful mediation of cytotoxicity by the drug
(Fig. 4C). The down-regulation effect was uMUC-1 specific because
housekeeping genes [18S (quantitative RT-PCR) and tubulin
(Western blot)] were not down-regulated (Fig. 4A and B).
Figure 4. Effects of doxorubicin on
uMUC-1 availability in BT-20 cells—in vitro
tests. A, quantitative RT-PCR (qRT-PCR )
shows a significant down-regulation of
uMUC-1 induced by doxorubicin
(P = 0.028). This effect is MUC-1
specific because 18S RNA is not
down-regulated (inset ). B, Western blot
shows a noticeable down-regulation of
uMUC-1 but not of tubulin. Arrow, MUC-1
band. C, TUNEL assay indicates that
doxorubicin treatment of BT-20 cells
induced apoptosis efficiently.
www.aacrjournals.org
OF5
Cancer Res 2009; 69: (3). February 1, 2009
Cancer Research
Figure 5. MN-EPPT uptake by BT-20 cells—in vitro studies. A, in vitro MRI of BT-20 cells treated with doxorubicin and further incubated with either MN-EPPT or
MN-SCR. B, the T2 relaxation times of cells incubated with MN-EPPT were significantly shorter than those of cells incubated with MN-SCR (P < 0.0001, n = 6).
T2 relaxation times of MN-EPPT–incubated cells increased following treatment with doxorubicin, as compared with PBS-treated controls (P = 0.0003, n = 6). In cells
incubated with MN-SCR, T2 relaxation times stayed unchanged. C, flow cytometry of BT-20 cells treated with doxorubicin and incubated with either MN-EPPT or
MN-SCR. D, the relative fluorescence of cells incubated with MN-EPPT was significantly higher than that of cells incubated with MN-SCR (P = 0.007). The relative
fluorescence of MN-EPPT–incubated cells was lower following treatment with doxorubicin, as compared with PBS-treated controls (P = 0.007). In cells incubated with
MN-SCR, T2 relaxation times stayed unchanged.
Discussion
reflecting the reduced availability of the antigen in the presence of
doxorubicin. By contrast, when MN-SCR was used as a contrast
agent, relative fluorescence levels in doxorubicin-treated cells were
not different than in controls (Fig. 5C and D). These studies
confirmed that MN-EPPT specific accumulation in breast adenocarcinoma cells is a function of the relative availability of the target
antigen.
These experiments established an in vitro framework for our
subsequent in vivo studies aiming to quantitatively assess relative
uMUC-1 availability in preclinical breast cancer tumors. To
accomplish this final goal, we analyzed tumor T2 relaxation times
of animals treated with doxorubicin or saline as described in the
previous section and calculated tumor delta-T2 values before the
beginning and after the completion of treatment. In animals
treated with doxorubicin, there was a 25% decrease in delta-T2
after treatment with doxorubicin (P = 0.04, n = 5; Table 1A). In
control mice treated with saline, there was no significant difference
between pretreatment and posttreatment delta-T2 (Table 1A). The
observed drop in delta-T2 after treatment with doxorubicin was
indicative of reduced relative MN-EPPT accumulation in the
tumors and was accompanied by a corresponding down-regulation
of uMUC-1 as seen by quantitative RT-PCR (P = 0.03; Table 1B).
Unlike in mice injected with the targeted MN-EPPT probe, in
control mice injected with MN-SCR as a contrast agent, there were
no significant differences in delta-T2 between mice treated with
doxorubicin and saline (results not shown). These studies
corroborated our in vitro findings and further confirmed that
MN-EPPT can be used to noninvasively assess relative uMUC-1
availability in breast tumors.
Cancer Res 2009; 69: (3). February 1, 2009
The recent past has witnessed considerable progress in the way
human tumors are characterized with a particular gain in knowledge
of cancer at the molecular level. This has resulted in a shift toward
using molecularly targeted therapies for cancer, necessitating the
development of corresponding molecular tools to determine which
patients are most likely to benefit from particular therapies (26).
Consequently, the capacity to noninvasively assess tumor anatomy
and physiology and at the same time registering aberrations in the
molecular phenotype of cells is essential for early diagnosis and
effective treatment. With these studies, we have shown the utility of
MRI for multiparametric characterization of tumor response to
chemotherapy in a preclinical breast cancer model. We have
confirmed that MRI and NIRF can be combined to obtain
complementary information about anatomic changes in tumor size
following treatment using the breast cancer model. In addition, we
have established that our contrast agent, by virtue of being target
specific for a tumor antigen, can provide molecular information
about target antigen expression when combined with MRI as a
modality with a high quantitative sensitivity to local contrast agent
abundance. This approach represents an example of noninvasive
semiquantitative assessment of therapy-induced changes in tumor
molecular expression levels by MRI.
An important point relevant to our studies is the potential for
effects other than probe uptake by the tumor, such as vascularity
and the presence of necrosis, which can affect the measurements
by MRI. Areas of active angiogenesis as well as edema can result in
elevated T2 relaxation times, whereas areas of local hemorrhage or
necrosis can appear with short T2. We have attempted to eliminate
OF6
www.aacrjournals.org
Monitoring Tumor Response to Chemotherapy
Table 1. MN-EPPT uptake by BT-20 tumors as a function of uMUC-1 expression—in vivo studies
(A) delta-T2 (MRI)
DOX
Saline
Pretreatment
Posttreatment
Mean (SE)
Mean (SE)
20.08107 (2.542284)
20.66182 (3.541263)
13.52276 (0.858644)
25.93164 (5.242064)
(B) uMUC-1 relative expression (qRT-PCR)
Mean (SE)
DOX
Saline
0.41 (0.19)
1.00 (0.14)
NOTE: A, average delta-T2 values (T2 post-contrast minus T2 pre-contrast, from T2 maps) of mice treated with doxorubicin or saline before and after
treatment, and using MN-EPPT as a contrast agent. There was a significant 25% (P = 0.04, n = 5) decrease in average tumor delta-T2 relaxation times
after treatment with doxorubicin. In saline-treated controls, delta-T2 values before and after treatment were not significantly different. B, quantitative
RT-PCR of BT-20 tumors from mice treated with doxorubicin and saline-treated controls. uMUC-1 expression levels in animals treated with doxorubicin
were significantly lower than in controls (P = 0.03).
Abbreviations: DOX, doxorubicin; qRT-PCR, quantitative RT-PCR.
similar probes can be developed for the monitoring of the
availability of antigens such as Her2/neu, epidermal growth factor
receptor, estrogen receptor, and somatostatin receptor, to name a
few. Progress in that direction has already been made (27–31). The
ultimate implications of the approach include optimization of
existing cancer treatment regimens, testing of novel therapeutic
paradigms, and development of personalized medicine protocols.
or minimize the influence of these factors by focusing on the deltaT2 parameter, which represents an estimate of the change in T2
after probe injection relative to preinjection. Assuming a minimal
change in anatomic/physiologic tumor characteristics during the
time between the preinjection and postinjection magnetic
resonance measurements, the delta-T2 value reflects relative probe
accumulation. Consequently, the development of similar targeted
probes with a faster clearance rate (e.g., Gd chelates) and optimal
target binding kinetics will permit a further shortening of the
timespan between pre-contrast and post-contrast imaging and will
even more accurately reflect probe uptake by the tumor.
In summary, this study describes a method for antigen-specific
noninvasive breast tumor imaging by MRI. Our strategy can be
used in vivo to monitor tumor response to therapy based not only
on changes in tumor size, which is a late and often unreliable
marker of response to therapy, but also on the expression of
molecular biomarkers. Considering that many newly established
and experimental breast cancer therapies are molecularly targeted,
References
1. Moore A, Medarova Z, Potthast A, Dai G. In vivo
targeting of underglycosylated MUC-1 tumor antigen
using a multimodal imaging probe. Cancer Res 2004;64:
1821–7.
2. Massoud TF, Gambhir SS. Molecular imaging in living
subjects: seeing fundamental biological processes in a
new light. Genes Dev 2003;17:545–80.
3. Medarova Z, Pham W, Kim Y, Dai G, Moore A. In vivo
imaging of tumor response to therapy using a dualmodality imaging strategy. Int J Cancer 2006;118:
2796–802.
4. Ollivier L, Balu-Maestro C, Leclere J. Imaging in
evaluation of response to neoadjuvant breast cancer
treatment. Cancer Imaging 2005;5:27–31.
www.aacrjournals.org
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
Acknowledgments
Received 5/28/2008; revised 9/25/2008; accepted 11/11/2008.
Grant support: NIH grant EB001727 (A. Moore).
The costs of publication of this article were defrayed in part by the payment of page
charges. This article must therefore be hereby marked advertisement in accordance
with 18 U.S.C. Section 1734 solely to indicate this fact.
We thank John Moore for excellent technical support in animal surgery.
5. Balu-Maestro C, Chapellier C, Bleuse A, Chanalet I,
Chauvel C, Largillier R. Imaging in evaluation of
response to neoadjuvant breast cancer treatment
benefits of MRI. Breast Cancer Res Treat 2002;72:
145–52.
6. Bruneton JN, Caramella E, Hery M, Aubanel D,
Manzino JJ, Picard JL. Axillary lymph node metastases
in breast cancer: preoperative detection with US.
Radiology 1986;158:325–6.
7. Yang WT, Ahuja A, Tang A, Suen M, King W, Metreweli
C. High resolution sonographic detection of axillary
lymph node metastases in breast cancer. J Ultrasound
Med 1996;15:241–6.
8. Schnall MD. Application of magnetic resonance
imaging to early detection of breast cancer. Breast
Cancer Res 2001;3:17–21.
9. Rieber A, Zeitler H, Rosenthal H, et al. MRI of breast
cancer: influence of chemotherapy on sensitivity. Br J
Radiol 1997;70:452–8.
10. Lee CH, Weinreb JC. The use of magnetic resonance
imaging in breast cancer screening. J Am Coll Radiol
2004;1:176–82.
11. Boetes C, Strijk SP, Holland R, Barentsz JO, Van Der
Sluis RF, Ruijs JH. False-negative MR imaging of
malignant breast tumors. Eur Radiol 1997;7:1231–4.
12. Mankoff DA, Dunnwald LK, Gralow JR, Ellis GK,
Drucker MJ, Livingston RB. Monitoring the response of
patients with locally advanced breast carcinoma to
neoadjuvant chemotherapy using [technetium 99m]sestamibi scintimammography. Cancer 1999;85:2410–23.
13. Schelling M, Avril N, Nahrig J, et al. Positron emission
tomography using [ 1 8 F]fluorodeoxyglucose for
OF7
Cancer Res 2009; 69: (3). February 1, 2009
Cancer Research
monitoring primary chemotherapy in breast cancer.
J Clin Oncol 2000;18:1689–95.
14. Benard F, Turcotte E. Imaging in breast cancer:
single-photon computed tomography and positronemission tomography. Breast Cancer Res 2005;7:153–62.
15. Apostolopoulos V, McKenzie I. Cellular mucins:
targets for immunotherapy. Crit Rev Immunol 1994;14:
293–309.
16. Harisinghani M, Barentsz J, Hahn P, et al. Noninvasive
detection of clinically occult lymph-node metastases in
prostate cancer. New Engl J Med 2003;348:2491–9.
17. Trail PA, Willner D, Lasch SJ, et al. Antigen-specific
activity of carcinoma-reactive BR64-doxorubicin conjugates evaluated in vitro and in human tumor
xenograft models. Cancer Res 1992;52:5693–700.
18. Groth-Pedersen L, Ostenfeld MS, Hoyer-Hansen M,
Nylandsted J, Jaattela M. Vincristine induces dramatic
lysosomal changes and sensitizes cancer cells to
lysosome-destabilizing siramesine. Cancer Res 2007;67:
2217–25.
19. Yang SZ, Kohno N, Kondo K, et al. Adriamycin
activates E-cadherin-mediated cell-cell adhesion in
human breast cancer cells. Int J Oncol 1999;15:1109–15.
20. Tanimoto A, Oshio K, Suematsu M, Pouliquen D,
Stark DD. Relaxation effects of clustered particles.
J Magn Reson Imaging 2001;14:72–7.
21. Molina R, Barak V, van Dalen A, et al. Tumor markers
in breast cancer—European Group on Tumor Markers
recommendations. Tumour Biol 2005;26:281–93.
22. Greenberg R, Barnea Y, Kaplan O, Kashtan H,
Skornick Y. Detection of cancer cells in the axillary
drainage using RT-PCR after operations for breast
cancer. Breast 2004;13:49–55.
23. Kontani K, Taguchi O, Narita T, et al. Modulation of
MUC1 mucin as an escape mechanism of breast cancer
cells from autologous cytotoxic T-lymphocytes. Br J
Cancer 2001;84:1258–64.
24. Al-azawi D, Kelly G, Myers E, et al. CA 15–3 is
predictive of response and disease recurrence following
treatment in locally advanced breast cancer. BMC
Cancer 2006;6:220.
25. Eom YW, Kim MA, Park SS, et al. Two distinct
modes of cell death induced by doxorubicin: apoptosis
and cell death through mitotic catastrophe accompanied by senescence-like phenotype. Oncogene 2005;24:
4765–77.
Cancer Res 2009; 69: (3). February 1, 2009
OF8
26. Sawyers CL. The cancer biomarker problem. Nature
2008;452:548–52.
27. Artemov D, Mori N, Okollie B, Bhujwalla ZM. MR
molecular imaging of the Her-2/neu receptor in breast
cancer cells using targeted iron oxide nanoparticles.
Magn Reson Med 2003;49:403–8.
28. Artemov D, Mori N, Ravi R, Bhujwalla ZM. Magnetic
resonance molecular imaging of the HER-2/neu receptor. Cancer Res 2003;63:2723–7.
29. Stefflova K, Li H, Chen J, Zheng G. Peptide-based
pharmacomodulation of a cancer-targeted optical
imaging and photodynamic therapy agent. Bioconjug
Chem 2007;18:379–88.
30. Neeman M, Degani H. Metabolic studies of estrogenand tamoxifen-treated human breast cancer cells by
nuclear magnetic resonance spectroscopy. Cancer Res
1989;49:589–94.
31. Smith-Jones PM, Stolz B, Bruns C, et al. Gallium67/gallium-68-[DFO]-octreotide—a potential radiopharmaceutical for PET imaging of somatostatin
receptor-positive tumors: synthesis and radiolabeling
in vitro and preliminary in vivo studies. J Nucl Med
1994;35:317–25.
www.aacrjournals.org