Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print February 27, 2004, 10.1148/radiol.2311021559
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2311021559v1
231/1/135    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
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 Luciani, A.
Right arrow Articles by Cuenod, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Luciani, A.
Right arrow Articles by Cuenod, C. A.
(Radiology 2004;231:135-142.)
© RSNA, 2004

Experimental Studies

Glucose-Receptor MR Imaging of Tumors: Study in Mice with PEGylated Paramagnetic Niosomes1

Alain Luciani, MD, Jean-Christophe Olivier, PharmD, PhD, Olivier Clement, MD, PhD, Nathalie Siauve, MD, PhD, Pierre-Yves Brillet, MD, Bertrand Bessoud, MD, Florence Gazeau, PhD, Ijeoma F. Uchegbu, PhD, Edmond Kahn, PhD, Guy Frija, MD and Charles A. Cuenod, MD, PhD

1 From the Radiology Department, Hôpital Européen Georges Pompidou, INSERM U494, LRI, Faculté Necker, 20 Rue Leblanc, 75015 Paris, France (A.L., O.C., N.S., P.Y.B., B.B., E.K., G.F., C.A.C.); Laboratoire de Pharmacie Galénique et Biopharmacie, Equipe émergente "Médicaments anti-infectieux et barrière Hémato-encéphalique," Faculté de Médecine et de Pharmacie, Poitiers, France (J.C.O.); CNRS UMR 7603, Laboratoire des Milieux Désordonnés Hétérogènes et de Pharmacologie Clinique LMDH, Faculté Paris VI-VII, France (F.G.); and Department of Pharmaceutical Sciences, University of Strathclyde, Strathclyde Institute for Biomedical Sciences, Scotland (I.F.U.). From the 2002 RSNA scientific assembly. Received November 25, 2002; revision requested February 20, 2003; revision received May 29; accepted June 30. Address correspondence to C.A.C. (e-mail: ca@cuenod.net).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
PURPOSE: To evaluate a magnetic resonance (MR) imaging contrast agent for tumor detection based on paramagnetic nonionic vesicles (niosomes) bearing polyethylene glycol (PEG) and glucose conjugates for the targeting of overexpressed glucose receptors.

MATERIALS AND METHODS: Four gadobenate dimeglumine–loaded niosome preparations including nonconjugated niosomes, niosomes bearing glucose conjugates (N-palmitoyl glucosamine [NPG]), niosomes bearing PEG 4400, and niosomes bearing both PEG and NPG were tested. In vitro cellular uptake was measured at electron paramagnetic resonance (EPR) after incubation with human prostate carcinoma, PC3, cells. In vivo distribution was studied at MR imaging 6, 12, and 24 hours after injection, with assessment of tumor, brain, liver, and muscle signal intensity (SI) in 49 mice bearing PC3 cells. Efficiency of targeted contrast agents was assessed with tumor-to-muscle contrast-to-noise ratio (CNR). Testing for differences was performed with analysis of variance followed by a posteriori Fisher test.

RESULTS: In vitro, gadolinium could be detected at EPR only in cell pellets incubated with niosomes bearing glucose conjugates or niosomes bearing both glucose conjugates and PEG (4.9 · 10−15 and 4.5 · 10−15 mol gadolinium per PC3 cell). In vivo, marked predominant tumor enhancement was demonstrated 24 hours after injection of glycosylated PEG niosomes (P < .01); no significant differences were observed following injection of nonconjugated niosomes, glycosylated niosomes, or PEG 4400 niosomes. Twenty-four hours after injection, sole presence of NPG or PEG 4400 on the surface of the niosome led to higher tumor-to-muscle CNR than that observed after injection of nonconjugated niosomes (CNR of 3.3 ± 0.7 [SD], 3.4 ± 2.2, and 0 ± 1.9). Combination of NPG and PEG led to even higher tumor-to-muscle CNR (6.3 ± 2.2).

CONCLUSION: Combination of PEG and glucose conjugates on the surface of niosomes significantly improved tumor targeting of an encapsulated paramagnetic agent assessed with MR imaging in a human carcinoma xenograft model.

© RSNA, 2004

Index terms: Animals • Glucose • Liposomes • Magnetic resonance (MR), contrast media • Neoplasms, experimental studies


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
The aim of imaging contrast agents is to improve detection or characterization of tissue lesions. Four classes of magnetic resonance (MR) contrast agents can be distinguished, and these include extracellular gadolinium chelates, blood pool agents, liver-targeted contrast agents, and tumor-targeted contrast agents (1). This latter class includes several preparations such as MR imaging contrast agents bound to vectors—such as antibodies, peptides, and polysaccharides—or entrapped within liposomes (2). Vectors are defined as inert carriers with a high affinity for a specific target. A variety of vectors have already been synthesized, and most investigators report the use of liposome-type vectors.

Liposomes are lipid vesicles that result from the combination of phospholipids or polar amphiphilic molecules in an aqueous medium (3–5). Gadolinium-labeled liposomes were initially developed as targeted MR contrast agents for the liver and spleen (6–8). Rapid blood clearance, however, is triggered by their interaction with opsonins and results inreticuloendothelial system uptake. In order to counter this rapid blood clearance after intravenous administration and degradation by molecular exchange with high-density lipoproteins (5), several researchers have used polyethylene glycol (PEG) as an external protective "coating" agent (9,10). PEG-coated liposomes have been named "stealth" liposomes because of their ability to limit reticuloendothelial system capture, thus leading to an increased plasma half-life, which in turn could favor tumor accumulation (11–18). The diminished plasma clearance that leads to an increased time of vesicle recirculation could be responsible for the passive targeting process (19,20).

However, while this passive targeting is well documented, active targeting of different phospholipid vesicles to tumor tissues or through the blood-brain barrier has not been studied as much as has passive targeting. The exploitation of the overexpression of glucose transporters, such as the glucose receptor GLUT-1, in tumor tissues and on the blood-brain barrier has already been successfully carried out in positron emission tomographic (PET) studies with fluorodeoxyglucose (21–23). Targeting of these overexpressed receptors by using glucose-type conjugates combined with lipid vesicles could thus favor tumor distribution in addition to the passive targeting process. Furthermore, the vesicular nature of such a new vector can enable the simultaneous entrapment of therapeutic agents (18,24,25). Targeting of both therapeutic and contrast agents to tumor cells could thus ensure an in vivo follow-up of the drug-targeting efficiency.

To date, as far as we know, no MR contrast agent that uses the overexpressed glucose receptors in tumor cells has been synthesized. We previously developed nonionic vesicles (niosomes) for drug-targeted delivery to glucose receptors by using N-palmitoyl glucosamine (NPG) as a glucose conjugate (26). Thus, the purpose of our study was to design and evaluate an MR imaging contrast agent for tumor detection based on entrapment of a paramagnetic agent within nonionic vesicles (niosomes) bearing both PEG and glucose conjugates for the targeting of overexpressed tumor glucose receptors.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Niosome Preparation and Characterization
Materials.—The chemical compounds used for niosome synthesis were palmitic acid N-hydroxysuccinimide (Sigma, Saint Quentin Fallavier, France), glucosamine (Sigma), sorbitan monostearate (Span 60; Sigma), cholesterol (Sigma), triethanolamine (Sigma), phosphate-buffered saline (pH 7.4) tablets (Sigma), chloroform (Sigma), isopropanol (Sigma), dimethyl sulfoxide and diethyl ether (Sigma), a mixture of cholesteryl poly-24-oxyethylene ether and cetyl poly-24-oxyethylene ether (Solulan C-24; Unipex, Rueil-Malmaison, France), stearyl poly-100-oxyethylene ether (Brij 700; Quimasso, Lyon, France), and the lipophilic gadolinium chelate gadobenate dimeglumine (Multihance; Bracco, Milan, Italy).

N-palmitoyl glucosamine synthesis.—The glucose conjugate N-palmitoyl glucosamine (NPG) was synthesized according to a procedure adapted from Dufes et al (26). Glucosamine (86.3 mg) was dissolved in dimethyl sulfoxide (15 mL) and triethanolamine (93 µL). To this, palmitic acid N-hydroxysuccinimide (283 mg) dissolved in chloroform (4 mL) was added. The mixture was stirred at room temperature for 48 hours and was protected from light. Chloroform was evaporated off at 40°C under low pressure. NPG was then precipitated with 45 mL purified water and was recovered on a sintered glass filter. The resulting powder was purified by washing consecutively with water (50 mL) and chloroform (50 mL) and was dried at 40°C.

Niosome preparation and size assessment.—Four formulations of gadobenate dimeglumine–loaded niosomes were prepared and characterized by an author (J.C.O.) (Table 1) by using the procedure adapted from Dufes et al (26). These formulations were nonconjugated niosomes, niosomes bearing a glucose conjugate (ie, NPG) that were named glycosylated niosomes, niosomes bearing a PEG 4400 coating that were named PEGylated niosomes, and niosomes bearing both NPG and PEG coating that were named glycosylated PEG 4400 niosomes. Figure 1 shows a schematic diagram of a glycosylated PEG 4400 niosome.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Niosome Formulations

 


View larger version (29K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1. Schematic representation of a glycosylated PEG 4400 niosome. NPG radicals (glucose conjugates) are included in the membrane of the vesicle. Gadobenate dimeglumine is entrapped within the vesicle. Span 60 is the trade name for sorbitan monostearate; Solulan C-24, trade name for a mixture of cholesteryl poly-24-oxyethylene ether and cetyl poly-24-oxyethylene ether; Brij 700, trade name for stearyl poly-100-oxyethylene ether.

 
Surfactants and cholesterol were dispersed in 1.5 mL of 0.5 mol/L gadobenate dimeglumine solution and 1.5 mL purified water. The mixture was heated in a water bath at 90°C for 30 minutes and was probe sonicated (Soniprep 150; Fisher Scientific, Loughborough, England) for 5 minutes with the amplitude set at 10 µm.

The mean niosome diameter was determined by using photon correlation spectroscopy (Zetasizer 5000; Malvern, Orsay, France).

Quantification of gadobenate dimeglumine incorporation with reversed-phase high-performance liquid chromatography and relaxometry.—To determine the percentage of gadobenate dimeglumine entrapment in niosomes, the preparations (500 µL) were diluted with 4 mL of a 2.6% sodium chloride solution, then ultracentrifuged (Optima L70 Preparative Ultracentrifuge with Ti. 70.1 rotor; Beckman Coulter, Fullerton, Calif) at 57,000 g and 4°C for 30 minutes. Supernatants were collected to determine free gadobenate dimeglumine with reversed-phase high-performance liquid chromatography (chelate determination) and relaxometry (Gd3+ determination).

The chromatographic system for reversed-phase high-performance liquid chromatography with UV detection of the chelate molecule (benzyloxypropionictetraacetate) included a pump (LC-10AT; Shimadzu, Kyoto, Japan), an autosampler (717 Plus; Waters, Milford, Mass), and a spectrophotometer (SpectroMonitor III MaxN series; Mipton Roy, Ivyland, Pa). Benzyloxypropionictetraacetate was analyzed with a 257-nm detection wavelength. Data were processed with an integrator (SP4290; Spectra-Physics, Mountain View, Calif). Percentage of entrapment, that is, the percentage of gadobenate dimeglumine entrapped within niosomes, was expressed as the percentage of the total amount (in moles) of gadobenate dimeglumine present in the preparation. Percentages of entrapment were determined for each batch of niosome preparations used in the study.

The T1 relaxation times of the niosome preparations in water were measured by two authors (A.L. and P.Y.B.) with a relaxometer (Minispec PC120; Bruker, Wissembourg, France) at 37°C with a 0.47-T magnet and an inversion-recovery sequence. With the determined T1 values for each sample and the concentration of gadobenate dimeglumine, or C, contained in each niosome preparation, we calculated the r1 of all preparations with Equation (1) as follows:

where T1m is the measured T1 and T1w is the T1 for water.

Gadolinium Uptake in Cultured Tumor Cells
To assess the influence of the niosome composition on tumor cell uptake, we measured the fraction of gadolinium taken up by cultured human prostate carcinoma, PC3, cells at electron paramagnetic resonance (EPR). Cells were cultured to confluence in 225 cm2 cell culture flasks (Falcon; Becton Dickinson, Heidelberg, Germany) at 37°C and 5% CO2 with medium (MEM{alpha}; Gibco BRL, Paisley, Scotland) supplemented with 10% fetal calf serum, 100 U/mL penicillin (Gibco BRL), and 100 µg/mL streptomycin (Gibco BRL). The culture medium was then removed, and each niosome preparation was added to achieve a final gadolinium concentration of 10 mmol/L. After incubation for 3 hours, the cells were washed twice with phosphate-buffered saline (Gibco BRL), recovered by means of trypsinization, and counted. After two cycles of centrifugation and redispersion in phosphate-buffered saline, the cell pellets were collected in quartz tubes and dehydrated at 80°C for 1 hour. The fraction of gadolinium taken up by the tumor cells was quantified by using pulsed EPR. The EPR static magnetic field was fixed at 9.25 GHz. EPR signal was converted to gadolinium concentration per cell after calibration with preparations of known gadolinium concentrations. All cell incubation procedures and gadolinium uptake measurements were each performed four times by two authors (A.L. and F.G.).

In Vivo MR Imaging
All animal experiments were performed in accordance with the National Institutes of Health recommendations, with induction of general anesthesia for every procedure.

A total number of 54 male nude (nu/nu) mice were received from a supplier (Iffa Credo, Paris, France). Anesthesia was induced with an intraperitoneal injection of ketamine hydrochloride (Imalgene; Merial, Lyon, France) and xylazine (Rompun; Bayer, Leverkusen, Germany). Of the total, 49 mice were included in the study because of deaths in three mice after anesthesia and in two mice after niosome injection.

Tumor xenograft.—Human prostate adenocarcinoma PC3 cells were implanted in mice by three authors (A.L., P.Y.B., and B.B.): One million cells were injected subcutaneously at the same levels on both sides of the animals at the thoracoabdominal junction level. Twenty-four mice developed two subcutaneous tumors, and 25 mice developed only one. A total of 73 tumors were therefore included in the study. Animals were injected with the contrast agent formulations for imaging when tumors were approximately 1 cm in diameter, that is, approximately 6 weeks after tumor cell injection.

Contrast agent administration.—The four gadobenate dimeglumine–based niosome preparations at a gadolinium concentration of 25 mmol/L were evaluated for MR imaging. For comparison, a solution of gadobenate dimeglumine was tested at the same concentration. The dose of gadolinium injected into mice through a tail vein was identical for all preparations and corresponded to an injection volume of 8 µL/g. Three intervals (6, 12, and 24 hours) between injection and MR imaging were studied. Three mice were studied per contrast agent (five types of contrast agents) and per interval, and four control mice were injected with 8 µL/g of isotonic saline solution. This method accounted for a total of 49 animals.

MR imaging and image analysis.—MR imaging was performed with four mice at a time and a 1.5-T unit (Signa; GE Medical Systems, Milwaukee, Wis) within a knee coil. For all imaging procedures, each animal that received general anesthesia was further immobilized in an individual plastic tube closely fitted around its body. The study was performed in the transverse plane with a T1-weighted spin-echo sequence (repetition time msec/echo time msec, 500/10; section thickness, 3 mm; field of view, 8 cm2; matrix, 256 x 192; number of signals acquired, three) and a total duration of 6 minutes 32 seconds. The animals were randomly assigned to both the type of contrast agent injected and the delay between injection and MR imaging.

To assess the in vivo distribution of the contrast agents, the signal intensity (SI) of brain, tumor, muscle, and liver was measured in regions of interest positioned by one author (A.L.) with free software (NIH Image; National Institutes of Health, Bethesda, Md) available at rsb.info.nih.gov/nih-image/. The mean region-of-interest sizes (mean number of pixels) for tumor, liver, brain, and muscle were 136, 175, 126, and 105, respectively. For all tumor SI measurements, the whole visible tumor was included in the region of interest. In order to avoid possible nonlinearly varying factors that could have influenced the SI measurements, a cylindric oil recipient was simultaneously imaged as an external standard, and SI values of all organs were divided by those of this external standard to obtain a relative signal intensity (RI) value. The mean RI for each organ in the four control mice without contrast injection, or RI0, was determined. The enhancement values, or ENH—expressed as percentages for each organ, contrast agent, and time lapse after injection—were calculated with Equation (2) as follows:

The contrast-to-noise ratio (CNR) was determined for tumor tissue as compared with muscle with Equation (3) as follows:

where SIorg is the SI of the organ studied, SImus is the SI of the muscle, and {Delta}N is the noise SD determined in the air above the mice.

Plasma pharmacokinetics.—Blood samples were drawn with cardiac puncture and a 26-gauge needle 6 hours after each contrast agent injection. Plasma was obtained following centrifugation, and gadolinium plasma concentrations were determined by using atomic emission spectroscopy.

Statistical Analysis
Testing for a difference in RI of the brain, the liver, the muscle, and the tumor after contrast medium injection within the different groups of mice was performed with analysis of variance followed by the a posteriori Fisher test. Significance level was set at .05. All statistical analyses were performed with software (Statview 4.5; Abacus Concepts, Berkeley, Calif) implemented on a computer (Macintosh; Apple, Palo Alto, Calif).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 
Characterization of Niosome Preparations
Mean diameter, percentage of gadobenate dimeglumine entrapment, and r1 are summarized in Table 2. Niosome diameters were similar for all preparations and ranged from 163 to 268 nm. The percentages of entrapped gadobenate dimeglumine were also similar for all preparations.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Characteristics of Formulations of Contrast Agent

 
Gadolinium Uptake in Cultured Tumor Cells
Gadolinium was detected with EPR only in cell pellets incubated with either glycosylated niosomes or glycosylated PEG 4400 niosomes. The EPR signal corresponded to a mean concentration of 4.9 · 10−15 mol gadolinium per PC3 cell (SD, 2 · 10−16) after cell incubation with glycosylated niosomes and of 4.5 · 10−15 mol gadolinium per PC3 cell (SD, 3 · 10−16) after cell incubation with glycosylated PEG 4400 niosomes. The EPR signal was too low to be detected within cells incubated with niosomes, PEG 4400 niosomes, or gadobenate dimeglumine solution.

In Vivo MR Imaging
Biodistribution of contrast agents.—The respective enhancement values of the tumor, brain, liver, and muscle after injection of each niosome preparation are shown in Figure 2.



View larger version (56K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2. Graphs show tumor ({blacktriangleup}), liver ({diamondsuit}), brain (x), and muscle ({bullet}) enhancement (mean ± standard error of the mean) after contrast agent injection. A, Nonconjugated niosomes. B, Glycosylated niosomes. C, PEG 4400 niosomes. D, Glycosylated PEG 4400 niosomes. Predominant tumor enhancement is demonstrated 24 hours after glycosylated PEG 4400 niosome injection.

 
Marked predominant tumor enhancement was demonstrated 24 hours after the injection of the glycosylated PEG 4400 niosomes (P < .01) (Fig 2,D), whereas no significant difference of tissue enhancement was observed after the injection of nonconjugated niosomes (Fig 2, A), glycosylated niosomes (Fig 2, B), or PEG 4400 niosomes (Fig 2, C). Predominant hepatic enhancement was observed after gadobenate dimeglumine injection.

Organ distribution.—The RI-to-time profiles for tumor, liver, and brain after injection of each contrast agent are shown in Figure 3.



View larger version (21K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3. Graphs show RI-to-time profiles (mean ± standard error of the mean) after injection of each contrast agent. A, Tumor. B, Liver. C, Brain. Highest tumor RI is observed 24 hours after glycosylated PEG 4400 niosome injection ({blacktriangleup}); highest liver RI is observed after gadobenate meglumine injection (x). In brain, RI-to-time profiles show a regular increase after glycosylated niosome ({blacksquare}) and glycosylated PEG 4400 niosome injection. {diamondsuit} = nonconjugated niosomes, {blacktriangledown} = PEG 4400 niosomes.

 
Tumor distribution is shown in Figure 3, A. Injection of glycosylated PEG 4400 niosomes resulted in a sustained increase of tumor RI values over time. Twenty-four hours after injection, the tumor RI values were significantly higher than those observed with any other contrast agents (P < .01, Fisher test) (Fig 4).



View larger version (95K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a. Transverse spin-echo T1-weighted MR images (500/10; section thickness, 3 mm) in two mice with xenograft PC3 tumors (arrowheads) 24 hours after injection of contrast agents. (a) Saline solution. (b) Glycosylated PEG 4400 niosomes. Image shows increased tumor RI and tumor-to-muscle CNR, with liver (Li) and gallbladder (Gb) enhancement. Lu = lung, m = muscle, * = heart.

 


View larger version (115K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b. Transverse spin-echo T1-weighted MR images (500/10; section thickness, 3 mm) in two mice with xenograft PC3 tumors (arrowheads) 24 hours after injection of contrast agents. (a) Saline solution. (b) Glycosylated PEG 4400 niosomes. Image shows increased tumor RI and tumor-to-muscle CNR, with liver (Li) and gallbladder (Gb) enhancement. Lu = lung, m = muscle, * = heart.

 
The maximum tumor RI after injection of PEG 4400 niosomes without NPG was observed at the 12 hours. At all times, the tumor RI values were inferior to those observed after injection of glycosylated PEG 4400 niosomes.

The maximum tumor RI after injection of glycosylated niosomes was observed at 6 hours, followed by a rapid decrease between 6 and 12 hours. Twenty-four hours after injection, the tumor RI was significantly higher than that observed after injection of nonconjugated niosomes (P = .002).

The injection of gadobenate dimeglumine resulted in a maximum tumor RI at 6 hours, followed by a rapid decrease. The injection of nonconjugated niosomes led to the lowest tumor RI at both 6 and 24 hours, with a progressive decrease in tumor RI between these two times.

Liver distribution is shown in Figure 3, B. The highest liver RI values were observed after injection of the liver-specific contrast agent gadobenate dimeglumine (Fig 3, B), with a maximum at 6 hours. No significant difference between the niosome preparations was observed regardless of the time considered.

Brain distribution is shown in Figure 3, C. In the brain, RI-to-time profiles for both glycosylated niosomes and glycosylated PEG 4400 niosomes showed a regular increase, with the highest values obtained 24 hours after injection, although no significant difference was demonstrated.

Muscle distribution was also evaluated. No significant difference in muscle RI was demonstrated regardless of the formulation used or the time at which the sample was studied.

CNR values.—The respective values of CNR for tumor versus muscle are summarized in Table 3. Regarding the type of contrast agent used, a significant difference for CNR was found only at 24 hours (P < .01). The highest tumor-to-muscle CNR was obtained after glycosylated PEG 4400 niosome injection. At all times considered, tumor CNRs after glycosylated niosome injection were systematically higher than those observed after nonconjugated niosome injection.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Tumor-to-Muscle CNR for Contrast Agents

 
Plasma Pharmacokinetics
The gadolinium concentrations in plasma measured 6 hours after injection by using atomic emission spectroscopy remained much higher with the PEG 4400-coated vesicles (57.5 µmol/L gadolinium for glycosylated PEG 4400 niosomes and 18.2 µmol/L for PEG 4400 niosomes) than with the non–PEG-coated vesicles (1.7 µmol/L gadolinium for glycosylated niosomes and 1.0 µmol/L for nonconjugated niosomes) or with gadobenate dimeglumine (0.22 µmol/L).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
REFERENCES
 
The detection of primary tumors, metastases, or local recurrences is a critical issue in oncology. We have developed a tumor-oriented MR contrast agent based on PEGylated niosome vesicles bearing NPG ligands targeted to the glucose receptors overexpressed in tumor cells. Our results suggest that this design allows a significantly higher in vivo tumor enhancement than that with nonconjugated niosomes in a human prostate carcinoma xenograft model.

The use of liposomes as drug carriers is well documented (27). It is known that pegylation of the liposomes improves their tumor accumulation by passive targeting. By limiting reticuloendothelial system uptake, the PEG coating results in prolonging the plasma half-life of the vesicles and thus in increasing their accumulation in tumor tissues through highly permeative neoangiogenic tumor capillaries (9,11,17–20,28,29). Similarly, we prepared PEG-coated niosomes. The advantages of using niosomes over liposomes are their improved in vitro stability and easier storage capabilities (26). Addition of PEG 4400 chains to the niosome surface did not significantly alter the percentage of entrapped gadobenate dimeglumine or its relaxivity. Our study findings confirm that surface PEG moieties of molecular weights beyond 2,000 Da favor tumor accumulation (9,11,14,17,18,30). Coating of niosomes with PEG 4400 moieties resulted in a diminished diameter of the vesicles that in itself may favor extracapillary penetration. However, a reduced diameter is also known to increase r1 (31), as our study results confirmed, and could in part have been responsible for the increased RI values observed with PEGylated niosomes compared with nonconjugated niosomes. However, significant differences in RI values between the niosome formulations were observed only in tumor tissue. Moreover, the tumor RI values were similar for both PEGylated and nonPEGylated niosomes 6 hours after injection. Therefore, differences in relaxivity values between the types of niosomes used did not bias our results. Our pharmacokinetic results 6 hours after injection further confirm that PEGylated niosomes enjoy an increased plasma half-life when compared with nonconjugated niosomes. A more extensive pharmacokinetic study will be required to confirm these findings.

The EPR measurements of gadolinium uptake in cultured PC3 cells demonstrate that the PEG coating alone is unable to achieve active tumor cell targeting. Thus, an additional tumor ligand proved mandatory. A variety of ligands have been proposed for improved tumor targeting, and these include antibodies, folates, biotin, peptides, and polysaccharides (32–47). Most literature data focus on immunoliposomes, that is, liposomes conjugated to antibodies against antigens expressed by tumor cells. The drawbacks of such an approach are the limitation of the targeting to a single tumor type and the potential induction of immune responses that could limit repeated administration. The strategy of using nonspecific folate, biotin, or polysaccharide ligands to target receptors overexpressed by a wide range of tumor cells is therefore more promising. Glucose receptors such as GLUT are overexpressed in many tumor cells because of their enhanced metabolism (48–50). This is extensively exploited by fluorodeoxyglucose PET for tumor imaging, as it is believed that glucose interaction with glucose receptors, and namely GLUT-1–type receptors, could be the rate-limiting step in fluorodeoxyglucose capture by the tumor cells (51,52). We included a glucose conjugate, the NPG surfactant, in niosomes to act as a tumor-targeting ligand. The incorporation of NPG in vesicles does not alter the stability of the vesicle or the relaxivity of the entrapped paramagnetic agent and improves tumor targeting. Glycosylated niosomes and glycosylated PEG 4400 niosomes were the only vesicles detected at EPR in PC3 tumor cells. Moreover, in vivo 24 hours after glycosylated PEG 4400 niosome injection, not only was the tumor RI significantly higher than that observed with the other preparations but the tumor-to-muscle CNR was maximum, which suggested a preferential tumor capture mediated by the NPG radical. On the opposite end of the spectrum, gadobenate dimeglumine injection led to a high tumor and liver RI 6 hours after injection but to a lower tumor-to-muscle CNR, and these results confirmed its hepatospecific distribution.

The in vivo effect of NPG radicals was further highlighted by the analysis of the brain RI variations after contrast agent injection. In the brain, continuous capillaries of the blood-brain barrier do not favor passive targeting, which could partly explain the relatively weak enhancement observed after PEG 4400 niosome injection. On the opposite end of the spectrum, the increase of brain RI over time observed after injection of either glycosylated niosomes or glycosylated PEG 4400 niosomes could be related to the interaction of NPG with the highly expressed glucose receptors on the blood-brain barrier (53). This feature is also exploited in functional PET imaging of the brain.

Niosome membranes are made of a bilayer of randomly distributed synthetic surfactants and cholesterol within and between the inner and the outer layers. Hydrophilic moieties or polar groups are turned toward the aqueous inner phase (niosome core) or to the outer phase (dispersion medium). The presence of NPG on the exterior niosome surface was previously demonstrated at transmission electron microscopy with concanavalin A lectin–gold complex (20 nm) conjugate as a specific marker (26). However, one limitation of our study was that we were unable to assess the precise position of the NPG radical on the niosome with regard to the position of the long chains of PEG 4400. This could have been a rate-limiting step in targeting efficiency, as NPG radicals might have been periodically shielded by the PEG chains. However, the fact that injection of only glycosylated niosomes or glycosylated PEG 4400 niosomes showed that cellular uptake at EPR confirms that NPG radicals were available on the outer surface of the niosomes to target GLUT receptors. Development of PEG chains with NPG radicals positioned at their extremity in order to improve availability of the glucose conjugate is currently being undertaken at Laboratoire de Pharmacie Galénique et Biopharmacie, Poitiers, France.

Our findings are preliminary. The efficiency of the targeting and the contrast modulation on the images should be enhanced to increase tumor detection. Gadobenate dimeglumine was selected as the contrast agent because of its lipophilic moieties that facilitate its link to the membrane surface of lipid vesicles (54). Other lipophilic contrast agents could be tested to improve the percentage of entrapment and the relaxivity of our contrast agent (55). The vesicular structure of our niosomes, on the other hand, enables the use of iron oxide particles with higher T2 and T2* relaxivities, which could further improve MR detection (13,56).

Assessment of SI with regions of interest that encompass the entire tumor could have contributed to underestimation of tumor enhancement after injection of contrast agents, as most of the enhancement was found in the peripheral part of the tumor. This measurement method is, however, easy to standardize, thereby avoiding an arbitrary and hardly reproducible selection of the tumor periphery in these less than 1-cm-wide xenograft tumors.

Some questions remain: We are currently unable to assess the exact location of gadobenate dimeglumine relative to niosomes (54). Furthermore, the mechanisms by which the niosomes that bear NPG ligands are associated to the tumor cells are still unknown. These mechanisms could be the result of a simple accumulation of vesicles at the surface of the tumor cell membranes driven by the NPG-GLUT ligand receptor association, of the incorporation of the lipid bilayers of the vesicles into the cell membranes, or of a phagocytic entrapment of the vesicles within the tumor cells with endosomal formation following the endocytosis pathway of GLUT receptors (49). Further experiments will be required to identify the involved mechanisms. Additional studies with alternate tumor cells and contrast agents are mandatory. Moreover, although liposomes are known as safe carriers, studies of toxic reactions are required in order to confirm the potential use of niosomes as targeting contrast agents in humans.

Practical application: Because of a simultaneous passive and active targeting, the combination of PEG radicals and glucose conjugates on the surface of niosomes ensured the tumor accumulation of an entrapped lipophilic paramagnetic agent that can be detected with MR imaging. The use of such vesicles for tumor detection takes advantage of the high spatial resolution of MR imaging and does not require the combination of two imaging techniques, such as PET and computed tomography (CT), for detection and localization. In addition to cost, one of the drawbacks of using two separate PET and CT acquisitions is the difficulty of image registration and fusion.

Furthermore, the vesicular structure of niosomes allows the simultaneous entrapment of both contrast agents and active drugs (18,24,25), thus enabling an in vivo follow-up of the drug distribution with MR imaging (56).


   ACKNOWLEDGMENTS
 
The authors thank the staff members of the MRI research center CIERM, Kremlin-Bicètre, France, and Jacques Bittoun, MD, PhD, for their help during MR imaging procedures. The authors acknowledge the help of Marie F. Poupon, MD, Institut Curie, Paris, France, for providing the PC3 prostate carcinoma cell lines. The authors also thank Joël Poupon, MD, Hôpital Lariboisière, Paris, France, for performing the atomic emission spectroscopic assays on the plasma samples and Michael Schwarzinger, MD, for having reviewed our manuscript with regard to statistical analysis.


   FOOTNOTES
 
See also Science to Practice in this issue.

Abbreviations: CNR = contrast-to-noise ratio, EPR = electron paramagnetic resonance, NPG = N-palmitoyl glucosamine, PEG = polyethylene glycol, RI = relative signal intensity, SI = signal intensity

Author contributions: Guarantors of integrity of entire study, C.A.C., O.C., G.F.; study concepts, C.A.C., A.L., J.C.O.; study design, C.A.C., O.C., N.S., A.L.; literature research, A.L., C.A.C., O.C.; experimental studies, A.L., F.G., B.B., P.Y.B., E.K., I.F.U., N.S.; data acquisition, A.L., P.Y.B., B.B.; data analysis/interpretation, A.L., C.A.C., J.C.O.; statistical analysis, A.L.; manuscript preparation, A.L., C.A.C., J.C.O., O.C.; manuscript definition of intellectual content, C.A.C., A.L., G.F., O.C.; manuscript editing, A.L., C.A.C.; manuscript revision/review, A.L., C.A.C., J.C.O., O.C., G.F.; manuscript final version approval, A.L., C.A.C., O.C.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
 REFERENCES
 

  1. Gupta H, Weissleder R. Targeted contrast agents in MR imaging. Magn Reson Imaging Clin N Am 1996; 4:171-184.[Medline]
  2. Weissleder R, Bogdanov A, Papisov M. Drug targeting in magnetic resonance imaging. Magn Reson Q 1992; 8:55-63.[Medline]
  3. Torchilin VP. Liposomes as delivery agents for medical imaging. Mol Med Today 1996; 2:242-249.[CrossRef][Medline]
  4. Unger EC, Shen DK, Fritz TA. Status of liposomes as MR contrast agents. J Magn Reson Imaging 1993; 3:195-198.[Medline]
  5. Gregoriadis G, Florence AT. Liposomes in drug delivery: clinical, diagnostic and ophthalmic potential. Drugs 1993; 45:15-28.
  6. Kabalka G, Buonocore E, Hubner K, Moss T, Norley N, Huang L. Gadolinium-labeled liposomes: targeted MR contrast agents for the liver and spleen. Radiology 1987; 163:255-258.[Abstract/Free Full Text]
  7. Unger EC, Winokur T, MacDougall P, et al. Hepatic metastases: liposomal Gd-DTPA-enhanced MR imaging. Radiology 1989; 171:81-85.[Abstract/Free Full Text]
  8. Seltzer SE, Blau M, Herman LW, et al. Contrast material–carrying liposomes: biodistribution, clearance, and imaging characteristics. Radiology 1995; 194:775-781.[Abstract/Free Full Text]
  9. Woodle MC, Lasic DD. Sterically stabilized liposomes. Biochim Biophys Acta 1992; 1113:171-199.[Medline]
  10. Allen TM, Hansen CB, Lopes de Menezes DE. Pharmacokinetics of long-circulating liposomes. Adv Drug Deliv Rev 1995; 16:267-284.[CrossRef]
  11. Papahadjopoulos D, Allen TM, Gabizon A, et al. Sterically stabilized liposomes: improvements in pharmacokinetics and antitumor therapeutic efficacy. Proc Natl Acad Sci U S A 1991; 88:11460-11464.[Abstract/Free Full Text]
  12. Nikolova AN, Jones MN. Effect of grafted PEG-2000 on the size and permeability of vesicles. Biochim Biophys Acta 1996; 1304:120-128.[Medline]
  13. Bulte JW, de Cuyper M, Despres D, Frank JA. Short- vs. long-circulating magnetoliposomes as bone marrow-seeking MR contrast agents. J Magn Reson Imaging 1999; 9:329-335.
  14. Hong RL, Huang CJ, Tseng YL, et al. Direct comparison of liposomal doxorubicin with or without polyethylene glycol coating in C-26 tumor-bearing mice: is surface coating with polyethylene glycol beneficial? Clin Cancer Res 1999; 5:3645-3652.[Abstract/Free Full Text]
  15. Nicholas AR, Scott MJ, Kennedy NI, Jones MN. Effect of grafted polyethylene glycol (PEG) on the size, encapsulation efficiency and permeability of vesicles. Biochim Biophys Acta 2000; 1463:167-178.[Medline]
  16. Harrington KJ, Rowlinson-Busza G, Syrigos KN, Uster PS, Abra RM, Stewart JS. Biodistribution and pharmacokinetics of 111In-DTPA-labelled pegylated liposomes in a human tumour xenograft model: implications for novel targeting strategies. Br J Cancer 2000; 83:232-238.[CrossRef][Medline]
  17. Harrington KJ, Rowlinson-Busza G, Syrigos KN, Uster PS, Vile RG, Stewart JS. Pegylated liposomes have potential as vehicles for intratumoral and subcutaneous drug delivery. Clin Cancer Res 2000; 6:2528-2537.[Abstract/Free Full Text]
  18. Harrington KJ, Rowlinson-Busza G, Syrigos KN, et al. Pegylated liposome-encapsulated doxorubicin and cisplatin enhance the effect of radiotherapy in a tumor xenograft model. Clin Cancer Res 2000; 6:4939-4949.[Abstract/Free Full Text]
  19. Wu NZ, Da D, Rudoll TL, Needham D, Whorton AR, Dewhirst MW. Increased microvascular permeability contributes to preferential accumulation of Stealth liposomes in tumor tissue. Cancer Res 1993; 53:3765-3770.[Abstract/Free Full Text]
  20. Needham D, McIntosh TJ, Lasic DD. Repulsive interactions and mechanical stability of polymer-grafted lipid membranes. Biochim Biophys Acta 1992; 1108:40-48.[Medline]
  21. Clavo AC, Brown RS, Wahl RL. Fluorodeoxyglucose uptake in human cancer cell lines is increased by hypoxia. J Nucl Med 1995; 36:1625-1632.[Abstract/Free Full Text]
  22. Haberkorn U, Ziegler SI, Oberdorfer F, et al. FDG uptake, tumor proliferation and expression of glycolysis associated genes in animal tumor models. Nucl Med Biol 1994; 21:827-834.[CrossRef][Medline]
  23. Maublant J, Vuillez JP, Talbot JN, et al. Positron emission tomography (PET) and (F-18)-fluorodeoxyglucose in (FDG) in cancerology. Bull Cancer 1998; 85:935-950. [French].[Medline]
  24. Monsky WL, Kruskal JB, Lukyanov AN, et al. Radio-frequency ablation increases intratumoral liposomal doxorubicin accumulation in a rat breast tumor model. Radiology 2002; 224:823-829.[Abstract/Free Full Text]
  25. Goldberg SN, Girnan GD, Lukyanov AN, et al. Percutaneous tumor ablation: increased necrosis with combined radio-frequency ablation and intravenous liposomal doxorubicin in a rat breast tumor model. Radiology 2002; 222:797-804.[Abstract/Free Full Text]
  26. Dufes C, Schatzlein AG, Tetley L, et al. Niosomes and polymeric chitosan based vesicles bearing transferrin and glucose ligands for drug targeting. Pharm Res 2000; 17:1250-1258.[CrossRef][Medline]
  27. Maruyama K, Ishida O, Takizawa T, Moribe K. Possibility of active targeting to tumor tissues with liposomes. Adv Drug Deliv Rev 1999; 40:89-102.[CrossRef][Medline]
  28. Yuan F, Leunig M, Huang SK, Berk DA, Papahadjopoulos D, Jain RK. Microvascular permeability and interstitial penetration of sterically stabilized (stealth) liposomes in a human tumor xenograft. Cancer Res 1994; 54:3352-3356.[Abstract/Free Full Text]
  29. La Noce A, Frigeni V, Filatori I, Danieli A, Tirone P. Neurotolerability of gadobenate dimeglumine: evaluation of brain dopamine concentration in freely moving rats by microdialysis. Invest Radiol 1999; 34:543-547.[CrossRef][Medline]
  30. Boucher Y, Leunig M, Jain RK. Tumor angiogenesis and interstitial hypertension. Cancer Res 1996; 56:4264-4266.[Abstract/Free Full Text]
  31. Fossheim SL, Fahlvik AK, Klaveness J, Muller RN. Paramagnetic liposomes as MRI contrast agents: influence of liposomal physicochemical properties on the in vitro relaxivity. Magn Reson Imaging 1999; 17:83-89.[CrossRef][Medline]
  32. Reimer P, Weissleder R, Brady TJ, et al. Experimental hepatocellular carcinoma: MR receptor imaging. Radiology 1991; 180:641-645.[Abstract/Free Full Text]
  33. Reimer P, Weissleder R, Lee AS, Buettner S, Wittenberg J, Brady TJ. Asialoglycoprotein receptor function in benign liver disease: evaluation with MR imaging. Radiology 1991; 178:769-774.[Abstract/Free Full Text]
  34. Blume G, Cevc G, Crommelin MD, Bakker-Woudenberg IA, Kluft C, Storm G. Specific targeting with poly(ethylene glycol)-modified liposomes: coupling of homing devices to the ends of the polymeric chains combines effective target binding with long circulation times. Biochim Biophys Acta 1993; 1149:180-184.[Medline]
  35. Eavarone DA, Yu X, Bellamkonda RV. Targeted drug delivery to C6 glioma by transferrin-coupled liposomes. J Biomed Mater Res 2000; 51:10-14.[CrossRef][Medline]
  36. Gabizon A, Horowitz AT, Goren D, et al. Targeting folate receptor with folate linked to extremities of poly(ethylene glycol)-grafted liposomes: in vitro studies. Bioconjug Chem 1999; 10:289-298.[CrossRef][Medline]
  37. Kresse M, Wagner S, Pfefferer D, Lawaczeck R, Elste V, Semmler W. Targeting of ultrasmall superparamagnetic iron oxide (USPIO) particles to tumor cells in vivo by using transferrin receptor pathways. Magn Reson Med 1998; 40:236-242.[Medline]
  38. Lee RJ, Low PS. Delivery of liposomes into cultured KB cells via folate receptor-mediated endocytosis. J Biol Chem 1994; 269:3198-3204.[Abstract/Free Full Text]
  39. Loughrey HC, Ferraretto A, Cannon AM, et al. Characterisation of biotinylated liposomes for in vivo targeting applications. FEBS Lett 1993; 332:183-188.[CrossRef][Medline]
  40. Maruyama K, Takizawa T, Takahashi N, Tagawa T, Nagaike K, Iwatsuru M. Targeting efficiency of PEG-immunoliposome-conjugated antibodies at PEG terminals. Adv Drug Deliv Rev 1997; 24:235-242.[CrossRef]
  41. Matsukawa S, Yamamoto M, Ichinose K, et al. Selective uptake by cancer cells of liposomes coated with polysaccharides bearing 1-aminolactose. Anticancer Res 2000; 20:2339-2344.[Medline]
  42. Nag A, Ghosh PC. Assessment of targeting potential of galactosylated and mannosylated sterically stabilized liposomes to different cell types of mouse liver. J Drug Target 1999; 6:427-438.[Medline]
  43. Sarti P, Ginobbi P, D’Agostino I, et al. Liposomal targeting of leukaemia HL60 cells induced by transferrin-receptor endocytosis. Biotechnol Appl Biochem 1996; 24:269-276.
  44. Singh M. Transferrin as a targeting ligand for liposomes and anticancer drugs. Curr Pharm Des 1999; 5:443-451.[Medline]
  45. Sunamoto J, Iwamoto K. Protein-coated and polysaccharide-coated liposomes as drug carriers. Crit Rev Ther Drug Carrier Syst 1986; 2:117-136.[Medline]
  46. Torchilin VP. Immunoliposomes and PEGylated immunoliposomes: possible use for targeted delivery of imaging agents. Immunomethods 1994; 4:244-258.[CrossRef][Medline]
  47. Yamamoto M, Ichinose K, Ishii N, et al. Utility of liposomes coated with polysaccharide bearing 1-amino-lactose as targeting chemotherapy for AH66 hepatoma cells. Oncol Rep 2000; 7:107-111.[Medline]
  48. Warburg O. On the origin of cancer cells. Science 1956; 309-314.
  49. Mueckler M. Facilitative glucose transporters. Eur J Biochem 1994; 219:713-725.[Medline]
  50. Hatanaka M. Transport of sugars in tumor cell membranes. Biochim Biophys Acta 1974; 355:77-104.[Medline]
  51. Chung JK, Lee YJ, Kim C, et al. Mechanisms related to [18F]fluorodeoxyglucose uptake of human colon cancers transplanted in nude mice. J Nucl Med 1999; 40:339-346.[Abstract/Free Full Text]
  52. Waki A, Kato H, Yano R, et al. The importance of glucose transport activity as the rate-limiting step of 2-deoxyglucose uptake in tumor cells in vitro. Nucl Med Biol 1998; 25:593-597.[CrossRef][Medline]
  53. Farrell CL, Pardridge WM. Blood-brain barrier glucose transporter is asymmetrically distributed on brain capillary endothelial lumenal and ablumenal membranes: an electron microscopic immunogold study. Proc Natl Acad Sci U S A 1991; 88:5779-5783.[Abstract/Free Full Text]
  54. Tilcock C, Ahkong QF, Koenig SH, Brown RD, Davis M, Kabalka G. The design of liposomal paramagnetic MR agents: effect of vesicle size upon the relaxivity of surface-incorporated lipophilic chelates. Magn Reson Med 1992; 27:44-51.[Medline]
  55. Schwendener RA, Wuthrich R, Duewell S, Wehrli E, von Schulthess GK. A pharmacokinetic and MRI study of unilamellar gadolinium-, manganese-, and iron-DTPA-stearate liposomes as organ-specific contrast agents. Invest Radiol 1990; 25:922-932.[CrossRef][Medline]
  56. Pauser S, Reszka R, Wagner S, Wolf KJ, Buhr HJ, Berger G. Liposome-encapsulated superparamagnetic iron oxide particles as markers in an MRI-guided search for tumor-specific drug carriers. Anticancer Drug Des 1997; 12:125-135.[Medline]



This article has been cited by other articles:


Home page
 RadiologyHome page
C. Riviere, M.-S. Martina, Y. Tomita, C. Wilhelm, A. Tran Dinh, C. Menager, E. Pinard, S. Lesieur, F. Gazeau, and J. Seylaz
Magnetic Targeting of Nanometric Magnetic Fluid loaded Liposomes to Specific Brain Intravascular Areas: A Dynamic Imaging Study in Mice
Radiology, August 1, 2007; 244(2): 439 - 448.
[Abstract] [Full Text] [PDF]

Home page
 RadiologyHome page
P. Reimer
Tumor-targeted MR Contrast Agents: Hype or Future Hope?
Radiology, April 1, 2004; 231(1): 1 - 2.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2311021559v1
231/1/135    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
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 Luciani, A.
Right arrow Articles by Cuenod, C. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Luciani, A.
Right arrow Articles by Cuenod, C. A.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE