Cancer Research Donn Young Jordan
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online
Cancer Research 68, 7676, September 15, 2008. doi: 10.1158/0008-5472.CAN-08-0689
© 2008 American Association for Cancer Research
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Oostendorp, M.
Right arrow Articles by Backes, W. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Oostendorp, M.
Right arrow Articles by Backes, W. H.

Systems Biology and Emerging Technologies

Quantitative Molecular Magnetic Resonance Imaging of Tumor Angiogenesis Using cNGR-Labeled Paramagnetic Quantum Dots

Marlies Oostendorp1,2, Kim Douma2,3, Tilman M. Hackeng2,4, Anouk Dirksen2,4, Mark J. Post2,3,5,7, Marc A.M.J. van Zandvoort2,3 and Walter H. Backes1,2,6

1 Department of Radiology, Maastricht University Hospital, 2 Cardiovascular Research Institute Maastricht (CARIM), Departments of 3 Biomedical Engineering, 4 Biochemistry, and 5 Physiology, and 6 Research Institute GROW, Maastricht University, Maastricht, the Netherlands; and 7 Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, the Netherlands

Requests for reprints: Walter H. Backes, Maastricht University Hospital, Department of Radiology, P.Debyelaan 25, 6229 HX, Maastricht, the Netherlands. Phone: 0031-43-3876948; Fax: 0031-43-3876909; E-mail: wbac{at}rdia.azm.nl.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 References
 
The objective of this study was to develop and apply cyclic Asn-Gly-Arg (cNGR)-labeled paramagnetic quantum dots (cNGR-pQDs) for the noninvasive assessment of tumor angiogenic activity using quantitative in vivo molecular magnetic resonance imaging (MRI). cNGR was previously shown to colocalize with CD13, an aminopeptidase that is highly overexpressed on angiogenic tumor endothelium. Because angiogenesis is important for tumor growth and metastatization, its in vivo detection and quantification may allow objective diagnosis of tumor status and evaluation of treatment response. I.v. injection of cNGR-pQDs in tumor-bearing mice resulted in increased quantitative contrast, comprising increased longitudinal relaxation rate and decreased proton visibility, in the tumor rim but not in tumor core or muscle tissue. This showed that cNGR-pQDs allow in vivo quantification and accurate localization of angiogenic activity. MRI results were validated using ex vivo two-photon laser scanning microscopy (TPLSM), which showed that cNGR-pQDs were primarily located on the surface of tumor endothelial cells and to a lesser extent in the vessel lumen. In contrast, unlabeled pQDs were not or only sparsely detected with both MRI and TPLSM, supporting a high specificity of cNGR-pQDs for angiogenic tumor vasculature. [Cancer Res 2008;68(18):7676–83]


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 References
 
Angiogenesis, the formation of new capillaries from existing blood vessels, is key to tumor growth and metastatization by providing proliferating tumor cells with oxygen and nutrients (1, 2). Moreover, angiogenic activity is related to tumor malignancy (3, 4). Noninvasive detection of angiogenic activity is therefore highly relevant for adequate tumor diagnosis. Quantification of angiogenesis may furthermore allow objective monitoring of tumor progression, for instance in response to treatment.

Currently, molecular imaging techniques are being developed that allow direct visualization and characterization of cellular or molecular activation of angiogenesis-related pathways (5). More specifically, molecular imaging uses contrast agents that home to up-regulated biomolecules (e.g., receptors, enzymes) via interaction with high-affinity ligands coupled to the contrast agent. Ideally, this results in an altered signal intensity at the location of these molecules. Of the different imaging modalities, magnetic resonance imaging (MRI) may be the most desirable for molecular imaging due to its excellent spatial resolution and soft tissue contrast. Moreover, molecular MRI potentially allows direct covisualization of tumor angiogenic activity with anatomy. However, the inherently low sensitivity of MRI is a problem due to the typically low abundance of up-regulated biomolecules. This can be overcome by large molecular weight constructs carrying a high payload of gadolinium or iron, and multiple targeting ligands to enhance the relaxivity and targeting efficacy, respectively, of the particle (6).

One of the best-defined ligands for molecular imaging of angiogenesis is the cyclic Arg-Gly-Asp (cRGD) peptide, which binds specifically to the {alpha}vβ3-integrin (7, 8). However, for the cyclic Asn-Gly-Arg (cNGR) motif, the tumor-homing capability was shown to be 3-fold higher compared with cRGD (9). The clinical applicability of cNGR as a tumor-homing ligand was previously shown by conjugating cNGR to tumor necrosis factor {alpha} (TNF{alpha}). Compared with unlabeled TNF{alpha}, cNGR-TNF{alpha} displayed a significantly increased antitumor activity with similar systemic toxicity (10–12).

The vascular address of cNGR is a specific isoform of CD13 (aminopeptidase N), a transmembrane glycoprotein involved in cancer angiogenesis, tumor invasion, and metastasis, which is overexpressed by activated endothelial cells (ECs) of tumor vasculature (9, 13, 14). CD13 is not required for vessel growth during embryonic development and normal adult function, as shown in CD13-null mice (15). In a model of retinal neovascularization, these mice had significantly decreased vessel growth, suggesting that CD13 is important in pathologic neovascularization. In addition, fluorophore-conjugated cNGR allowed detection of the in vivo expression of CD13 in tumors and infarcted myocardium (16, 17). Competition with unconjugated cNGR significantly decreased the fluorescence signal, indicating high specificity of cNGR for CD13 (16, 17).

Despite the aforementioned high tumor-homing capability of cNGR, its potency as a targeting ligand for molecular imaging of tumor angiogenesis is currently unknown. Therefore, the objective of this study was to explore cNGR-labeled paramagnetic quantum dots (cNGR-pQDs) for the noninvasive and selective in vivo detection of tumor neovascularization using quantitative molecular MRI. QDs were chosen as contrast agent scaffolds because of their excellent photophysical properties, i.e., broad excitation, small emission spectra, and limited photobleaching (18, 19). Furthermore, QDs enabled binding of multiple targeting ligands and gadolinium chelates. The bimodal nature of the particle (i.e., paramagnetic and fluorescent) allowed validation of the results with ex vivo two-photon laser scanning microscopy (TPLSM). With TPLSM, three-dimensional contrast agent localization can be obtained at subcellular resolution with a penetration depth reaching 250 µm in tumors.

MRI data were analyzed via absolute quantification of contrast agent induced changes in the longitudinal relaxation rate R1 (1/T1) of the tissue, which is proportional to contrast agent concentration, and proton visibility. The latter expectedly decreases at high densities of paramagnetic contrast material. Quantitative analysis requires acquisition of a series of images and may provide improved sensitivity of molecular MRI. Theoretically, the used inversion recovery (IR) technique has an inherent 2-fold higher sensitivity than spin echo pulse sequences and by measuring a series of images it potentially allows detection of smaller changes in R1 than a single image.

Both MRI and TPLSM showed specific binding of cNGR-pQDs to ECs in the angiogenic tumor rim but not in tumor core or muscle tissue. Furthermore, a significantly lower quantitative contrast was found with unlabeled pQDs, indicating a high specificity of the cNGR-labeled contrast agent for angiogenic ECs. To our knowledge, this is the first noninvasive in vivo application of cNGR as a targeting ligand for molecular MRI of tumor angiogenesis.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 Disclosure of Potential...
 References
 
Preparation of cNGR-Labeled Paramagnetic Quantum Dots
NAc-Cys(4MeBzl)-Asn(Xanthyl)-Gly-Arg(Tosyl)-Cys(4MeBzl)-Gly-Gly-Lys(Fmoc)-peptide was synthesized by tBoc solid phase peptide synthesis, as described previously (16, 20). On the resin, the lysine side chain was selectively deprotected by treatment with 20% piperidine/dimethylformamide (4 x 3 min). After covalent coupling of biotin-succinimidyl ester (Molecular Probes) to the lysine {varepsilon}-amino group to obtain biotinylated peptide-resin, the peptide was deprotected and cleaved from the resin using anhydrous hydrogen fluoride for 1 h at 0°C with 4% p-cresol as scavenger and lyophilized. Electrospray ionization mass spectrometry (ESI-MS) revealed a mass of 1,287.4, corresponding well to the theoretical average mass (1,288.7) of the reduced biotinylated Nac-Cys-Asn-Gly-Arg-Cys-Gly-Gly-Lys(biotin)-NH2 peptide. Oxidative folding of the crude product in 0.1 mol/L Tris (pH 8), 1 mol/L guanidin at 4°C for 16 h yielded the internal disulfide bridged biotin-cyclic NGR, which was high performance liquid chromatography (HPLC)-purified (C18 RP-HPLC) and lyophilized. ESI-MS confirmed a mass decrease of 2, representing the loss of 2 protons from the cystein side chains due to the generation of 1 disulfide bond (S-S). Biotinylated poly(lysine) dendritic wedge, a construct comprising 8 Gd-diethylenetriaminepentaacetic acid (DTPA) moieties, was synthesized and purified similarly (21, 22).

Curnis and colleagues (23) previously showed that cNGR spontaneously converts into isoDGR by asparagine deamidation at slightly basic pH, generating an {alpha}vβ3-integrin ligand. Using a combination of HPLC and mass spectrometry up to 24 h after dissolving cNGR in water (pH 6.0) and 1 µmol/L borate buffer (pH 8.3; supplemented with 0.05% NaN3), respectively, it was found that this process did not occur in the time-period of the experiments (data not shown).

Bimodal, multivalent contrast agent was prepared as follows. Streptavidin-conjugated QDs [1 µmol/L in borate buffer (pH 8.3), emission at 585 nm] were purchased from Invitrogen. QDs were composed of a CdSe core with a ZnS shell and covered with polyethyleneglycol-2000. Each QD holds ~10 surface-bound streptavidins, allowing 30 biotinylated compounds to bind on average.8 For each experiment, cNGR-pQDs were prepared freshly at room temperature by sequential mixing of 100 µL QD solution with biotin-cNGR and biotin-poly(lysine) dendritic wedge, both dissolved in HBSS (pH 7.4; Invitrogen), in a molar ratio of 1:6:24 to a total volume of ~120 µL. Samples were mildly vortexed during each preparation step to ensure a homogeneous distribution of biotin-cNGR and biotin-poly(lysine) dendritic wedge over the QD surface. Overall, each QD carried a maximum of 192 Gd ions and 6 cNGR peptides. Unlabeled pQDs carried the same number of Gd constructs but no cNGR. A schematic representation of the cNGR-pQD particle is shown in Fig. 1 .


Figure 1
View larger version (27K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Schematic representation of a cNGR-labeled paramagnetic quantum dot. Each particle carries ~10 streptavidin moieties to which 6 cNGR groups and 24 dendritic gadolinium constructs were bound. The total number of gadolinium ions per particle was maximally 192.

 
Animal Model
All animal studies were approved by the institutional animal welfare committee. Human colorectal adenocarcinoma cells (1.5 to 3·106; LS174T; American Type Culture Collection CL-188) were s.c. and unilaterally injected on the flank of ~15-wk-old male Swiss nu/nu mice (Charles River). Mice were subjected to the MRI examination when the tumor diameter was >1.0 cm, which was ~16 d after LS174T injection.

For in vivo MRI, mice were anesthetized using 1.5% to 2.0% isoflurane (Abbott Laboratories Ltd) in medical air and were placed prone in a dedicated animal holder with built-in mask for anesthesia gas supply. An infusion line was placed in the jugular vein for contrast agent administration during the MRI experiment. A heating pad was placed over the mice to maintain normothermic conditions. Respiration rate and body temperature were continuously monitored via a balloon sensor and rectal temperature probe, respectively, interfaced to an MR compatible small animal monitoring system (SA Instruments, Inc.).

Mice were randomly selected for injection with either cNGR-labeled or unlabeled pQDs. Seven mice were included for each contrast agent group. Mice were kept inside the magnet during the entire MRI experiment to preserve their position.

MRI Protocol
All MRI experiments were performed on a 7 T Bruker Biospec 70/30 USR MRI system, interfaced to an AVANCE II console (Bruker Biospin GmbH). The BGA12-S mini imaging gradient (maximum gradient strength, 720 mTm–1; slew rate, 6,000 Tm–1s–1) and a 3.5-cm inner diameter quadrature volume resonator were used.

Molecular MRI. Tumors were localized using T2-weighted anatomic images (TR, 4,200 ms; TE, 37.4 ms). Next, precontrast R1 values were determined using a series of IR measurements with increasing inversion times (TR, 4,000 ms; TE, 8.4 ms; TI, 500, 1,000, 1,500, 2,000, 2,500, and 3,500 ms; total scan time, 18 min). Subsequently, mice were injected with 120 µL of cNGR-labeled or unlabeled pQDs, followed by a 50 µL saline flush. IR experiments were repeated ~30 min postcontrast to ensure adequate contrast agent circulation and a reduced level of intravascular contrast agent. Images were recorded using a field of view (FOV) of 4.0 x 4.0 cm2, a 192 x 192 acquisition matrix interpolated to 256 x 256 by means of zero-filling, and a slice thickness of 1.2 mm, resulting in 0.16 x 0.16 x 1.2-mm3-sized voxels. On average, 15 contiguous slices were recorded in multislice mode (range, 11–22 slices; depending on tumor size and orientation). After MRI, mice were euthanized by cervical dislocation.

Competition experiment. Four tumor-bearing mice were randomly selected for a competition experiment of cNGR-pQDs with unconjugated cNGR, i.e., nonbiotinylated, nonparamagnetic, and nonfluorescent. Imaging was performed as described above, except that 525 µg per mouse of unconjugated cNGR, i.e., a 1,000-fold excess compared with QD-bound cNGR, was injected i.v. 10 min after administration of cNGR-pQDs.

Biodistribution. Healthy Swiss mice (Charles River) were injected with either cNGR-pQDs, unlabeled pQDs, or no contrast agent. After ~1 h circulation time, mice were sacrificed and whole body T1-weighted spin echo images were recorded (TR, 1,100 ms; TE, 8.5 ms; FOV, 4.0 x 6.0 cm2; matrix, 256 x 512; resolution, 0.16 x 0.12 x 1.2 mm3). Two mice were included per group.

Tissue harvesting. After MRI, tumor, spleen, liver, kidney, hind limb muscle, heart, and lung were excised and embedded in optimal cutting temperature (OCT) compound (Sakura Finetek Europe). Next, tissues were snapfrozen in cold 2-methylbutane (Acros Organics) for ~2 min and subsequently transferred to liquid nitrogen. Tissues were stored at –80°C until TPLSM measurements.

Contrast agent relaxivity. T1 relaxivity (r1) was determined by diluting cNGR-pQDs in HBSS in 9 steps to concentrations of 0 to 0.001 mmol/L (corresponding gadolinium concentrations, 0–0.192 mmol/L). The R1 of each sample was determined using the IR series as described above. Absolute gadolinium concentrations were measured using Inductively Coupled Plasma Mass Spectrometry. Longitudinal relaxivity was determined by the slope of a linear fit of R1 versus gadolinium concentration.

MRI Data Analysis
All data processing was performed in Matlab (The Mathworks), unless stated otherwise. IR images were first spatially coregistered using the mutual information algorithm in the MIRIT software package (24) to correct for possible animal motion in the images with different T1 contrast, and smoothed with a three dimensional Gaussian kernel with a full-width-at-half-maximum of 0.4 x 0.4 x 3.0 mm3. Regions of interest (ROI) were drawn manually in MRIcro (25) to define tumor and muscle tissue. Both T1- and T2-weighted images were used to accurately delineate tumors from surrounding tissue and edema.

Precontrast and postcontrast R1 values were determined on a voxel-by-voxel basis by nonlinear curve fitting of the IR signal intensity function (26):

Formula(1)
using the Levenberg-Marquardt optimization algorithm. S0 is a scaling factor including proton density, excitation pulse profile, echo time, and preamplifier gain.

The detection limit for changes in R1 ({Delta}R1 = R1,post – R1,pre) was determined with a Monte Carlo simulation using Eq. 1, in vivo relaxation rates, and representative noise levels as derived from the in vivo experiments. A voxel was considered significantly enhanced when {Delta}R1 was >1.96 (i.e., 95% confidence interval) times higher than the detection limit of 0.005 s–1. We defined the Quantitative Contrast derived from the {Delta}R1 measurements (QCR1) as the product of the mean {Delta}R1 and the percentage of significantly enhanced voxels for each tissue type, i.e., tumor rim and core, and muscle tissue. QCR1 indicates both the level and spatial extent of contrast agent binding. Changes in S0 ({Delta}S0 = S0,post – S0,pre) were also evaluated, and the Quantitative Contrast from S0 (QCS0) was defined analogously to QCR1 to yield a quantity that reflects proton visibility (27).

Tumor rim/core analysis. To investigate the differences between tumor rim, i.e., the region with the highest expected angiogenic activity, and core, the tumor rim was first defined as a ~1-mm thick peripheral zone with the strongest R1 enhancement, in accordance with the approach taken by others (28, 29). Using this thickness, the difference between cNGR-labeled and unlabeled pQDs was maximal (Fig. 3C). The rim comprised 29.0% ± 5.8% and 31.6% ± 3.5% of all tumor voxels for mice injected with cNGR-labeled and unlabeled pQDs, respectively. The tumor core was defined as the difference between whole tumor and tumor rim ROIs. Second, a contour was drawn to calculate the number of voxels with a significantly increased {Delta}R1 as a function of the distance to the tumor rim. As an empirical measure of spatial heterogeneity in angiogenic tumor activity, the half-value-depth was defined as the distance from the rim at which the percentage of enhanced voxels has decreased by 50% compared with its value at zero distance, i.e., the rim. The half-value-depth was calculated by fitting the group-averaged data presented in Fig. 3C with a monoexponential decay function.


Figure 3
View larger version (24K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Spatial distribution of angiogenic activity. A and B, Quantitative Contrast as derived from changes in R1 (QCR1; A) and from changes in S0 (QCS0; B) for tumor rim, tumor core, and hind limb muscle tissue. Data are shown for cNGR-pQDs (n = 7), unlabeled pQDs (n = 7), and the competition experiment of cNGR-pQDs with excess unconjugated cNGR (n = 4). C, percentage of enhanced voxels at a certain distance versus distance from the tumor rim for cNGR-labeled and unlabeled pQDs. Although enhanced voxels were mostly found at the tumor rim for both contrast agents, more than twice as many rim voxels were enhanced for cNGR-pQDs than for unlabeled pQDs. In the tumor core, similar values were found for both cNGR-labeled and unlabeled pQDs. Columns, median; bars, SE. *, P < 0.05; #, P < 0.05 for the O'Brien-Läuter summary measure of QCR1 and QCS0 (see text).

 
Biodistribution. ROIs defining the spleen, liver, kidney, heart, lung, and aorta were drawn manually in MRIcro. Signal intensities were averaged over the entire tissue and normalized to hind limb muscle.

Statistical analysis. Statistical analysis of paired samples was performed using a nonparametric Wilcoxon signed-rank test in SPSS 14.0 (SPSS).

As both QCR1 and QCS0 represent contrast agent presence, QCR1 and QCS0 were combined to a summary value according to O'Brien and Läuter (30, 31), which is more sensitive to contrast effects than the individual measures. Therefore, QCR1 and QCS0 were first standardized by z = [QC – mean(QC)/sd(QC)]. Subsequently, the absolute values of zQCR1 and zQCS0 were averaged per animal. The resulting summary measure was tested using a nonparametric Mann-Whitney U test. P < 0.05 was considered statistically significant.

TPLSM Data Acquisition
Tissue samples were thawed and washed with HBSS to remove OCT compound. Except for the spleen and liver, tissues were incubated with 25-fold diluted {alpha}CD31-FITC (0.5 mg/mL; BD Biosciences PharMingen) to fluorescently label ECs. Next, tissues were embedded in 2 w% agarose gel (Invitrogen), with their rim upwards. For measurements in the tumor core, tumors were cut transversally to resemble the slice orientation of the MRI measurements.

TPLSM imaging was performed using a Nikon Eclipse E600FN upright microscope, incorporated in the Bio-Rad Radiance 2100MP imaging system and operated by Lasersharp2000 V6.0 (Bio-Rad). Tissue samples were excited by the Tsunami Ti:sapphire laser (Spectra-Physics), which was pumped by a Millennia Vs 5 W pump laser (Spectra-Physics) and mode locked at 800 nm, with a 82.5 MHz repetition rate and 100 fs pulse width. Tissues were observed through a water dipping 60 x fluor objective with a 1.00 numerical aperture (Nikon). Photomultiplier tubes (PMTs 9108B02 and 9136B05; Electron Tubes Limited) were used to acquire fluorescence photons in three spectral regions: 420 to 470 nm (autofluorescence), 520 to 560 nm, (FITC) and 570 to 600 nm (QD). Each PMT was tuned for minimal bleed through of the fluorescent markers to adjacent PMTs. Images, color coded blue, green, and red, respectively, were subsequently merged into a single image. The in-plane pixel dwell time was 11.8 µs, which, together with a 2-fold Kalman averaging, resulted in an imaging speed of 0.16 Hz. The FOV was 179 x 179 µm2 with a matrix size of 512 x 512, resulting in 0.35 x 0.35 µm2 sized pixels.

TPLSM Data Analysis
Data were analyzed with Image-Pro Plus 6.0 (MediaCybernetics) and ImageJ 1.35 (NIH). Image quality was improved by convolution with a 1.05 x 1.05 µm2 Gaussian filter. Spatial distribution of pQDs was classified into four groups: intravascular, intracellular, colocalized with the EC membrane, or extravasated to the interstitium.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 Disclosure of Potential...
 References
 
In vivo targeting of activated tumor endothelium. The ability of cNGR to target angiogenic tumor ECs was evaluated in tumor-bearing nude mice by injecting them with cNGR-labeled or unlabeled pQDs. Tumor volumes of cNGR and control groups did not differ on MR images (mean ± SD, 1.0 ± 0.7 cm3 and 1.0 ± 0.6 cm3, respectively).

For both cNGR-labeled and unlabeled pQDs, changes in R1 ({Delta}R1) were spatially heterogeneous throughout the tumor and were most pronounced at the tumor rim (Fig. 2A ). Averaged over all mice, the {Delta}R1 induced by cNGR-pQDs ranged up to ~0.3 s–1, which was considerably larger than the intrinsic variation in precontrast tumor R1 of 0.1 s–1. Furthermore, the range in {Delta}R1 was relatively large compared with the precontrast tumor R1 of 0.8 s–1. Administration of unlabeled pQDs resulted in a 3-fold lower response range ({Delta}R1 < 0.1 s–1) compared with cNGR-pQDs.


Figure 2
View larger version (30K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. T2-weighted anatomic images with color overlay of {Delta}R1 (A) and {Delta}S0 (B) for tumor (T) and muscle (M) tissue of mice injected with cNGR-labeled or unlabeled pQDs (n = 7 for both groups). Changes in R1 were most pronounced at the tumor rim for cNGR-pQDs. Although an R1 increase in the tumor rim was also observed for unlabeled pQDs, the average response was 3-fold lower when compared with cNGR-pQDs, indicating a high specificity of cNGR for angiogenic tumor endothelium. This is further supported by the low changes in R1 found in muscle tissue. Changes in S0 (B) colocalized almost completely with changes in R1 (A). Representative TPLSM images of tumor (C) and muscle tissue (D) showing pQD signal (red) and EC-specific {alpha}CD31-FITC (green). cNGR-pQDs accurately colocalized with tumor ECs, indicating binding of the contrast agent to the tumor endothelium (C). cNGR-pQDs were also detected in muscle tissue with TPLSM (Fig. D, arrows), although to a much lesser extent than in tumor tissue. cNGR-pQDs did not display any colocalization with muscle ECs and were only found intraluminally. Unlabeled pQDs were not or only sparsely detected in both tumor and muscle tissue. Bar, 50 µm.

 
Subsequent investigation by TPLSM allowed localization of cNGR-labeled and unlabeled pQDs at a subcellular resolution. cNGR-pQDs were found to colocalize approximately thrice more often with tumor ECs than unlabeled pQDs (Fig. 2). cNGR-labeled and unlabeled pQDs were also found in the vessel lumen, albeit that cNGR-pQDs were approximately thrice more prevalent than unlabeled pQDs. Both contrast agents were only sparsely found to have extravasated into the tumor interstitium. Although cNGR was previously reported to be an internalizing peptide (32), cNGR-pQDs were not detected inside ECs with TPLSM.

Further evidence for the specificity of cNGR was provided by {Delta}R1 in hind limb muscle. Here, average {Delta}R1 upon administration of cNGR-pQDs was considerably lower than in the tumor and ranged up to 0.05 s–1. TPLSM did not display colocalization of cNGR-pQD with ECs of muscle vasculature. However, the incidence of cNGR-pQDs in the muscle vascular lumen was almost 2-fold higher than for unlabeled pQDs (Fig. 2D).

S0-effect. For both cNGR-labeled and unlabeled pQDs, changes in the scaling factor S0 colocalized strongly with {Delta}R1 (Fig. 2B). The S0-effect is likely caused by field inhomogeneities (T2*-effect) in the vicinity of the contrast agent, induced by the magnetic properties of QDs (33) and the dense gadolinium concentration on the particle. Analogous to iron oxide particles, such properties result in locally reduced transverse relaxation times T2 and T2*, a shift in local resonance frequency and a broader water resonance line, which is reflected by a decrease in S0, i.e., a reduced proton visibility (27, 34). Therefore, {Delta}R1 and {Delta}S0 both represent contrast agent presence.

Spatial heterogeneity. To explore the absolute differences between tumor rim, tumor core, and muscle, QCR1 and QCS0 were determined for each tissue type for cNGR-labeled and unlabeled pQDs (Fig. 3A and B ). Administration of cNGR-pQDs resulted in an ~50-fold increase in QCR1 in the angiogenic rim compared with tumor core or muscle tissue. For unlabeled pQDs, significant differences were also found between tumor rim and core, and tumor rim and muscle tissue, although the net increase in QCR1 was lower than for cNGR-pQDs. The decreases in S0 showed the same trend as the increases in R1 (cf. Fig. 3A and B).

For each of the three tissue types, no significant differences in QCR1 or QCS0 were found between cNGR-labeled and unlabeled pQDs. Because {Delta}R1 and {Delta}S0 were shown to accurately colocalize (Fig. 2), QCR1 data were combined with QCS0 to a summary measure as described above. This resulted in a statistically significant difference between cNGR-labeled and unlabeled pQDs for the tumor rim only (Fig. 3A and B).

To further investigate the spatial distribution of angiogenic activity in the tumor, the percentage of significantly enhanced voxels was calculated as a function of the distance to the tumor rim (Fig. 3C). Although the highest signal increase was found at the tumor rim for both contrast agents, more than twice as many rim voxels were enhanced for cNGR-pQDs than for unlabeled pQDs. In the tumor core, similar enhancements were found for both contrast agents. These findings qualitatively concur with previous findings showing that angiogenic activity is most pronounced at the tumor rim for this tumor model (7, 35).

Subsequently, half-value-depths were calculated for both cNGR-labeled and unlabeled pQDs. High values indicate a more homogeneous distribution of enhanced voxels over the entire tumor and thus a low spatial heterogeneity, whereas low values indicate a high spatial variation. For cNGR-labeled and unlabeled pQDs, the half-value-depths were 0.6 and 1.1 mm, respectively, indicating a stronger contrast between tumor rim and core for cNGR-pQDs, which suggests that cNGR-pQDs allow a better differentiation between tumor rim and core than unlabeled pQDs.

Competition experiment. I.v. injection of a 1,000-fold excess of unconjugated cNGR 10 min after administration of cNGR-pQDs resulted in a statistically significant decrease in QCR1 and QCS0 for the tumor rim (Fig. 3A and B). With TPLSM, cNGR-pQDs were barely detected in the tumor rim, which confirmed the MRI results (data not shown). These results therefore indicate that binding of cNGR-pQDs to tumor ECs is specific, reversible, and can be competed with unconjugated cNGR.

Biodistribution. Figure 4 shows the relative MRI signal intensities for the blood and major organs recorded ~1 hour after the administration of cNGR-pQDs, unlabeled pQDs, or no contrast agent. No differences were found between cNGR-labeled and unlabeled pQDs. Both contrast agents accumulated mainly in the spleen, liver, and kidneys (Fig. 4), which was confirmed by TPLSM and corresponds to previous findings (36). Due to the i.v. administration, pQDs were also expected to accumulate in the lung. However, MRI has only limited signal sensitivity in the lung due to the inherently low signal intensity and air-tissue interfaces. With TPLSM, pQDs could be clearly detected in the lung (Fig. 4), although microscopic imaging was hampered by tissue movement caused by heating of the sample during excitation, resulting in expansion of air in the pulmonary alveoli.


Figure 4
View larger version (39K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4. Biodistribution of cNGR-labeled and unlabeled pQDs (n = 2 for each group). A, postmortem T1-weighted MRI results. Signal intensities were normalized to hind limb muscle signal and subsequently averaged. No CA, no contrast agent administration (n = 2). Both cNGR-labeled and unlabeled pQDs were found to accumulate mainly in the spleen, liver, and kidneys. Columns, median; bars, SE. B, representative TPLSM images of spleen, liver, lung, and kidney. Because a similar biodistribution was found for cNGR-labeled and unlabeled pQDs, no differentiation was made for the TPLSM results. Red, pQDs; green, {alpha}CD31-FITC; blue, autofluorescence; bar, 50 µm.

 
Contrast agent relaxivity. The ionic T1 relaxivity of cNGR-pQDs, i.e., per Gd ion, was 7.1 ± 0.4 mmol/L–1s–1 at 7 T and 20°C, which lies in the expected range for macromolecular contrast agents and is in correspondence with previously reported values for Annexin A5 conjugated pQDs (37).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
Disclosure of Potential...
 References
 
Current findings. In this study, the ability of cNGR-labeled paramagnetic QDs to visualize and quantify angiogenic activity in LS174T tumors was evaluated using two complementary imaging modalities: in vivo MRI and ex vivo TPLSM. To our knowledge, this study shows the first results of the application of cNGR for molecular MRI. First, cNGR-pQDs were found to have a 3-fold higher quantitative MRI contrast in the tumor rim, i.e., the tumor region with the highest angiogenic activity, compared with unlabeled pQDs. Second, cNGR-pQDs were barely detectable in muscle tissue, indicating a high specificity of cNGR-pQDs for angiogenic vessels. Third, ex vivo TPLSM showed colocalization of cNGR-pQDs, but not unlabeled pQDs, with ECs in tumor vasculature. Fourth, cNGR-pQDs allowed a more accurate assessment of the rim-core heterogeneity in tumor angiogenic activity. Fifth, the competition experiment indicated that the binding of cNGR-pQDs to tumor ECs is specific, reversible, and can be competed with excess unconjugated cNGR. Finally, no qualitative differences in biodistribution were found between cNGR-labeled and unlabeled pQDs with MRI and TPLSM. For a full quantitative biodistribution assessment, Positron Emission Tomography would be better suited because the local relaxivity of the MRI contrast agent, which is required to convert R1 to concentrations, is unknown. Taken together, the presented results designate cNGR as an effective ligand for discriminating between quiescent and activated endothelium and for quantifying the extent of tumor angiogenic activity.

In correspondence with other studies (7, 35), our results indicate that angiogenic activity is highest at the tumor rim, which was reflected by a high QCR1 and a strongly negative QCS0 for cNGR- pQDs. This was supported by the statistically significant difference between cNGR-labeled and unlabeled pQDs found using the summary value. Unlabeled pQDs also showed differences between tumor rim versus core and muscle, albeit smaller than for the cNGR-pQDs. This is likely due to the intrinsically higher vascular density of the tumor rim and corresponding blood pool fraction, resulting in a larger amount of circulating contrast agent compared with tumor core or muscle tissue (35). Additionally, heterogeneous blood flow and long wash-in and wash-out times of macromolecular contrast agents, which was previously described for dendritic agents (38), may have contributed to a prolonged retention of unlabeled pQDs in the tumor peripheral vasculature.

Methodologic considerations. Nonlinear fitting of the IR signal over a range of inversion times provided a sensitive and reliable method for detecting R1 changes induced by contrast agent binding. Compared with signal intensity measurements, it is relatively independent of technical settings, e.g., repetition time, echo time, and flip angle, thereby allowing objective comparison between different subjects, both spatially and temporally. A disadvantage of the quantitative approach is the lengthy acquisition time because an adequate number of data points is required for accurate fitting of the IR curve. Although prolonged precontrast and postcontrast acquisition of a single image at a fixed inversion time may also allow accurate detection of changes in signal intensity, this will not provide quantitative information on R1 and S0. In our quantitative approach, coregistration of these variables ensured increased sensitivity to detect differences between cNGR-labeled and unlabeled pQDs and is therefore preferred over acquisition of a single image.

Theoretically, the measured longitudinal relaxation rates and contrast agent relaxivity allow estimation of local contrast agent concentration and might be used to gain insight in the density of the molecular target. However, the conditions under which the relaxivity was determined differ strongly from the in vivo situation. Not only is the relaxivity affected by the chemical environment of the particle, i.e., aqueous buffer versus blood plasma (39), specific binding to vascular endothelium reduces its tumbling rate, thereby increasing the relaxivity. Taken together, this would lead to an overestimation of the local concentration in vivo. Unfortunately, accurate measurements of in vivo relaxivity are currently unavailable.

Contrast agent. Quantum dots were chosen as scaffold to enable bimodal, i.e., MRI and TPLSM, visualization of tumor angiogenic activity, which is an essential step in the characterization and validation of cNGR as a targeting ligand. Furthermore, streptavidin-coated QDs provide a suitable and versatile research scaffold to identify and test other potential targeting ligands. In addition, QDs show minimal extravasation, both from healthy and hyperpermeable tumor vasculature,9 which is beneficial for EC targeting. However, QDs may pose serious health limitations due to the potential release of toxic cadmium ions. Although this can be prevented by effective shielding of the core (40, 41), QDs are not cleared from the body and accumulate in spleen, liver, and kidneys. Cadmium-based QDs will therefore not be approved for clinical application. Recently developed nontoxic and renally excretable QDs may provide a potential solution for this problem (42, 43). Nevertheless, once a robust MRI method has been accepted in clinical practice, validation with luminescent particles is no longer necessary and clinically more suitable particles may be applied.

The magnetic and semiconductive properties of QDs give increase to field inhomogeneities when placed inside a magnetic field (33), which likely result in a local decrease of the transverse relaxation times T2 and T2*. In addition, T2 contrast becomes more effective at high field strength, whereas T1 contrast decreases. Using a standard multislice multiecho spin echo sequence, an average R2 increase of 5.7 s–1 upon cNGR-pQD injection was detected in the tumor rim at a mean tumor precontrast R2 of ~27 s–1. However, the percentage of significantly enhanced voxels in the tumor rim was only 4%, which is considerably lower than the 42% found for R1. This shows that T2 changes did not interfere with the effects on T1 and S0. Consequently, the T1 and S0 quantification was more sensitive in discriminating between angiogenic activity in the tumor rim, tumor core, and muscle tissue than T2.

Clinical perspectives. Regarding the potential clinical applicability, quantitative molecular MRI with a suitable contrast agent has a number of advantages over the currently used immunohistochemical methods to quantify tumor angiogenic activity. First, molecular MRI is noninvasive and does not interfere with tissue integrity. Second, it can probe the entire tumor, whereas immunohistochemistry requires biopsies at one or multiple selected locations. Third, it allows covisualization of angiogenic activity with local anatomy. Fourth, tumor status or therapeutic response may be objectively monitored over time due to the absolute quantification methodology. Finally, molecular MRI allows direct detection of activated endothelium in functional vasculature, whereas immunohistochemistry measures both perfused and nonperfused vessels.

With respect to the applied tumor model, a human colorectal adenocarcinoma, MRI is clinically important for local T-staging of rectal cancer and for the identification of tumors close to or invading the mesorectal fascia (44). On diagnostic T2-weighted images, however, it remains difficult to differentiate between fibrotic tissue and viable tumor cells. Molecular MRI of angiogenesis may facilitate this demarcation because only viable tumor cells induce angiogenesis, which may be visualized upon administration of the targeted contrast agent.

Besides the availability of suitable contrast agents, clinical implementation of quantitative molecular MRI requires rapid imaging techniques. Possible sequences that allow fast quantification of relaxation times are Look-Locker (45), IR-true-FISP (46), and the recently described QRAPTEST (47). However, these methods are relatively sensitive to subject movement and field inhomogeneities, although the Look-Locker method was recently modified to allow in vivo T1-mapping of the heart (48). Thus, the development of fast quantification of relaxation rates seems to support future clinical application of quantitative molecular MRI.

In summary, we have shown that cNGR-labeled paramagnetic quantum dots are suitable for the noninvasive visualization and quantification of tumor angiogenic activity using in vivo molecular MRI. These results provide a promising basis for further developments in contrast agent design and synthesis, data acquisition, and postprocessing techniques, which may be valuable for future clinical applications to pathologies in which abnormal vessel growth plays a pivotal role.


   Disclosure of Potential Conflicts of Interest
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
References
 
No potential conflicts of interest were disclosed.


   Acknowledgments
 
Grant support: This work was sponsored by the SenterNovem BSIK program 'Molecular imaging of ischemic heart disease' (grant number 03033). The TPLSM was financed by the Dutch Scientific Organization (NWO 902-16-276).

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 Remco Foppen and Stephen H. Chamberlain from Invitrogen for providing detailed information on quantum dot composition; Sander Langereis, Jeannette Smulders, and Thea Haex from Philips Research, Eindhoven, the Netherlands for performing the ICP-MS measurements; and Ludwig Dubois for the in vitro growth of the LS174T tumor cell line.


   Footnotes
 
Note: M. Oostendorp and K. Douma contributed equally to this work.

8 Personal communication with Invitrogen. Back

9 Unpublished observations. Back

Received 2/25/08. Revised 6/12/08. Accepted 6/13/08.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 Disclosure of Potential...
 References
 

  1. Carmeliet P. Angiogenesis in health and disease. Nat Med 2003;9:653–60.[CrossRef][Medline]
  2. Folkman J. Tumor angiogenesis: therapeutic implications. N Engl J Med 1971;285:1182–6.[Medline]
  3. Meitar D, Crawford SE, Rademaker AW, Cohn SL. Tumor angiogenesis correlates with metastatic disease, N-myc amplification, and poor outcome in human neuroblastoma. J Clin Oncol 1996;14:405–14.[Abstract/Free Full Text]
  4. Daldrup H, Shames DM, Wendland M, et al. Correlation of dynamic contrast-enhanced MR imaging with histologic tumor grade: comparison of macromolecular and small-molecular contrast media. AJR Am J Roentgenol 1998;171:941–9.[Abstract/Free Full Text]
  5. Massoud TF, Gambhir SS. Molecular imaging in living subjects: seeing fundamental biological processes in a new light. Genes Dev 2003;17:545–80.[Free Full Text]
  6. Sipkins DA, Cheresh DA, Kazemi MR, Nevin LM, Bednarski MD, Li KC. Detection of tumor angiogenesis in vivo by {alpha}Vβ3-targeted magnetic resonance imaging. Nat Med 1998;4:623–6.[CrossRef][Medline]
  7. Mulder WJ, Strijkers GJ, Habets JW, et al. MR molecular imaging and fluorescence microscopy for identification of activated tumor endothelium using a bimodal lipidic nanoparticle. FASEB J 2005;19:2008–10.[Abstract/Free Full Text]
  8. Zhang C, Jugold M, Woenne EC, et al. Specific targeting of tumor angiogenesis by RGD-conjugated ultrasmall superparamagnetic iron oxide particles using a clinical 1.5-T magnetic resonance scanner. Cancer Res 2007;67:1555–62.[Abstract/Free Full Text]
  9. Arap W, Pasqualini R, Ruoslahti E. Cancer treatment by targeted drug delivery to tumor vasculature in a mouse model. Science 1998;279:377–80.[Abstract/Free Full Text]
  10. Curnis F, Sacchi A, Borgna L, Magni F, Gasparri A, Corti A. Enhancement of tumor necrosis factor {alpha} antitumor immunotherapeutic properties by targeted delivery to aminopeptidase N (CD13). Nat Biotechnol 2000;18:1185–90.[CrossRef][Medline]
  11. Zarovni N, Monaco L, Corti A. Inhibition of tumor growth by intramuscular injection of cDNA encoding tumor necrosis factor {alpha} coupled to NGR and RGD tumor-homing peptides. Hum Gene Ther 2004;15:373–82.[CrossRef][Medline]
  12. van Laarhoven HW, Gambarota G, Heerschap A, et al. Effects of the tumor vasculature targeting agent NGR-TNF on the tumor microenvironment in murine lymphomas. Invest New Drugs 2006;24:27–36.[CrossRef][Medline]
  13. Curnis F, Arrigoni G, Sacchi A, et al. Differential binding of drugs containing the NGR motif to CD13 isoforms in tumor vessels, epithelia, and myeloid cells. Cancer Res 2002;62:867–74.[Abstract/Free Full Text]
  14. Pasqualini R, Koivunen E, Kain R, et al. Aminopeptidase N is a receptor for tumor-homing peptides and a target for inhibiting angiogenesis. Cancer Res 2000;60:722–7.[Abstract/Free Full Text]
  15. Rangel R, Sun Y, Guzman-Rojas L, et al. Impaired angiogenesis in aminopeptidase N-null mice. Proc Natl Acad Sci U S A 2007;104:4588–93.[Abstract/Free Full Text]
  16. Buehler A, van Zandvoort MA, Stelt BJ, et al. cNGR: a novel homing sequence for CD13/APN targeted molecular imaging of murine cardiac angiogenesis in vivo. Arterioscler Thromb Vasc Biol 2006;26:2681–7.[Abstract/Free Full Text]
  17. von Wallbrunn A, Waldeck J, Holtke C, et al. In vivo optical imaging of CD13/APN-expression in tumor xenografts. J Biomed Opt 2008;13:011007.[CrossRef][Medline]
  18. Chan WC, Nie S. Quantum dot bioconjugates for ultrasensitive nonisotopic detection. Science 1998;281:2016–8.[Abstract/Free Full Text]
  19. Michalet X, Pinaud FF, Bentolila LA, et al. Quantum dots for live cells, in vivo imaging, and diagnostics. Science 2005;307:538–44.[Abstract/Free Full Text]
  20. Schnolzer M, Alewood P, Jones A, Alewood D, Kent SB. In situ neutralization in Boc-chemistry solid phase peptide synthesis. Rapid, high yield assembly of difficult sequences. Int J Pept Protein Res 1992;40:180–93.[Medline]
  21. Dirksen A, Langereis S, de Waal BFM, van Genderen MHP, Hackeng TM, Meijer EW. A supramolecular approach to multivalent target-specific MRI contrast agents for angiogenesis. Chem Commun 2005;2811–3. Epub 2005 Apr 14.
  22. Dirksen A, Meijer EW, Adriaens W, Hackeng TM. Strategy for the synthesis of multivalent peptide-based nonsymmetric dendrimers by native chemical ligation. Chem Commun (Camb) 2006;1667–9. Epub 2006 Mar 8.
  23. Curnis F, Longhi R, Crippa L, et al. Spontaneous formation of L-isoaspartate and gain of function in fibronectin. J Biol Chem 2006;281:36466–76.[Abstract/Free Full Text]
  24. Maes F, Collignon A, Vandermeulen D, Marchal G, Suetens P. Multimodality image registration by maximization of mutual information. IEEE Trans Med Imaging 1997;16:187–98.[CrossRef][Medline]
  25. Rorden C, Brett M. Stereotaxic display of brain lesions. Behav Neurol 2000;12:191–200.[Medline]
  26. Haacke EM, Brown RW, Thompson MR, Venkatesan R. Introductory Signal Acquisition Methods: Free Induction Decay, Spin Echoes, Inversion Recovery and Spectroscopy. Magnetic Resonance Imaging Physical Principles and Sequence Design. 1st ed. New York: John Wiley & Sons, Inc.; 1999. p. 111–38.
  27. Gambarota G, van Laarhoven HW, Philippens M, et al. Assessment of absolute blood volume in carcinoma by USPIO contrast-enhanced MRI. Magn Reson Imaging 2006;24:279–86.[CrossRef][Medline]
  28. de Lussanet QG, Beets-Tan RG, Backes WH, et al. Dynamic contrast-enhanced magnetic resonance imaging at 1.5 Tesla with gadopentetate dimeglumine to assess the angiostatic effects of anginex in mice. Eur J Cancer 2004;40:1262–8.[CrossRef][Medline]
  29. Padhani AR. Functional MRI for anticancer therapy assessment. Eur J Cancer 2002;38:2116–27.[CrossRef][Medline]
  30. O'Brien PC. Procedures for comparing samples with multiple endpoints. Biometrics 1984;40:1079–87.[CrossRef][Medline]
  31. Läuter J. Exact t and F tests for analyzing studies with multiple endpoints. Biometrics 1996;52:964–70.[CrossRef]
  32. Ellerby HM, Arap W, Ellerby LM, et al. Anti-cancer activity of targeted pro-apoptotic peptides. Nat Med 1999;5:1032–8.[CrossRef][Medline]
  33. Manninen A, Koskinen M, Reimann SM, Mottelson B. Magnetic properties of quantum dots and rings. Eur Phys J D 2001;16:381–5.[CrossRef]
  34. Wismer GL, Buxton RB, Rosen BR, et al. Susceptibility induced MR line broadening: applications to brain iron mapping. J Comput Assist Tomogr 1988;12:259–65.[Medline]
  35. de Lussanet QG, Backes WH, Griffioen AW, van Engelshoven JM, Beets-Tan RG. Gadopentetate dimeglumine versus ultrasmall superparamagnetic iron oxide for dynamic contrast-enhanced MR imaging of tumor angiogenesis in human colon carcinoma in mice. Radiology 2003;229:429–38.[Abstract/Free Full Text]
  36. Yang RS, Chang LW, Wu JP, et al. Persistent tissue kinetics and redistribution of nanoparticles, quantum dot 705, in mice: ICP-MS quantitative assessment. Environ Health Perspect 2007;115:1339–43.[Medline]
  37. Prinzen L, Miserus RJ, Dirksen A, et al. Optical and magnetic resonance imaging of cell death and platelet activation using annexin a5-functionalized quantum dots. Nano Lett 2007;7:93–100.[CrossRef][Medline]
  38. Langereis S, de Lussanet QG, van Genderen MHP, et al. Evaluation of Gd(III)DTPA-terminated poly(propylene imine) dendrimers as contrast agents for MR imaging. NMR Biomed 2006;19:133–41.[CrossRef][Medline]
  39. Stanisz GJ, Henkelman RM. Gd-DTPA relaxivity depends on macromolecular content. Magn Reson Med 2000;44:665–7.[CrossRef][Medline]
  40. Cho SJ, Maysinger D, Jain M, Roder B, Hackbarth S, Winnik FM. Long-term exposure to CdTe quantum dots causes functional impairments in live cells. Langmuir 2007;23:1974–80.[CrossRef][Medline]
  41. Yu WW, Chang E, Falkner JC, et al. Forming biocompatible and nonaggregated nanocrystals in water using amphiphilic polymers. J Am Chem Soc 2007;129:2871–9.[CrossRef][Medline]
  42. Soo Choi H, Liu W, Misra P, et al. Renal clearance of quantum dots. Nat Biotechnol 2007;25:1165–70.[CrossRef][Medline]
  43. Zimmer JP, Kim SW, Ohnishi S, Tanaka E, Frangioni JV, Bawendi MG. Size series of small indium arsenide-zinc selenide core-shell nanocrystals and their application to in vivo imaging. J Am Chem Soc 2006;128:2526–7.[CrossRef][Medline]
  44. Vliegen RF, Beets GL, von Meyenfeldt MF, et al. Rectal cancer: MR imaging in local staging-is gadolinium-based contrast material helpful? Radiology 2005;234:179–88.[Abstract/Free Full Text]
  45. Freeman A, Gowland P, Jellineck D, et al. Dynamic T1 studies of gadolinium uptake in brain tumors using LL-EPI. MAGMA 1994;2:409–12.[CrossRef]
  46. Schmitt P, Griswold MA, Jakob PM, et al. Inversion recovery TrueFISP: quantification of T([1]), T([2]), and spin density. Magn Reson Med 2004;51:661–7.[CrossRef][Medline]
  47. Warntjes JB, Dahlqvist O, Lundberg P. Novel method for rapid, simultaneous T1, T*2, and proton density quantification. Magn Reson Med 2007;57:528–37.[CrossRef][Medline]
  48. Messroghli DR, Radjenovic A, Kozerke S, Higgins DM, Sivananthan MU, Ridgway JP. Modified Look-Locker inversion recovery (MOLLI) for high-resolution T1 mapping of the heart. Magn Reson Med 2004;52:141–6.[CrossRef][Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
M. Oostendorp, M. J. Post, and W. H. Backes
Vessel Growth and Function: Depiction with Contrast-enhanced MR Imaging
Radiology, May 1, 2009; 251(2): 317 - 335.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Oostendorp, M.
Right arrow Articles by Backes, W. H.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Oostendorp, M.
Right arrow Articles by Backes, W. H.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Cancer Research Clinical Cancer Research
Cancer Epidemiology Biomarkers & Prevention Molecular Cancer Therapeutics
Molecular Cancer Research Cancer Prevention Research
Cancer Prevention Journals Portal Cancer Reviews Online
Annual Meeting Education Book Meeting Abstracts Online