点击显示 收起
1 From the Department of Radiology (M.T., J.S., S.W., D.K., J.H., B.H.) and Institute for Medical Biometry (T.S.), Charité, Medizinische Fakultät, Humboldt-Universität zu Berlin, Schumannstrasse 20/21, 10098 Berlin, Germany; Research Laboratory, Ferropharm, Teltow, Germany (S.W., H.P.); Institute for Veterinary Anatomy, Department of Veterinary Anatomy, Freie Universität Berlin, Germany (C.A., H.H.); and Department of Medical Engineering, Siemens Medical Systems, Erlangen, Germany (G.L.). Received August 29, 2000; revision requested October 11; final revision received July 25, 2001; accepted July 30. Supported by the Bundesministerium für Wirtschaft und Technologie as part of the FUTOUR program (project 03FU096B, reference F229). Address
| ABSTRACT |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
MATERIALS AND METHODS: The particles used in this study were coated with citrate as the monomer (VSOP-C91). The particles have a total diameter of 7 nm and show the following relaxivities at 0.47 T: T1, 19 L/mmol · sec-1; T2, 29 L/mmol · sec-1. Fifteen cardiac MR examinations were performed at 1.5 T in five pigs. Images were acquired from immediately to 35 minutes (equilibrium phase) after intravenous injection of gadopentetate dimeglumine, gadobenate dimeglumine, and the very small SPIO particles (n = 5 for each substance).
RESULTS: Immediately after administration of gadopentetate dimeglumine, gadobenate dimeglumine, and the very small SPIO particles, respectively, increases in the signal-to-noise ratio in blood were 94%, 103%, and 102% and in myocardium were 83%, 83%, and 29% (P < .05, very small SPIO particles versus the low–molecular-weight gadolinium-based compounds). Differences in the blood-to-myocardium contrast-to-noise ratio and visualization of the coronary arteries and their branches were also significant.
CONCLUSION: VSOP-C91 significantly improves visualization of the coronary arteries at MR angiography from immediately to 35 minutes after injection.
Index terms: Contrast media, experimental studies • Gadolinium • Magnetic resonance (MR), contrast media, 50.12143 • Magnetic resonance (MR), vascular studies, 50.12142, 50.12143
| INTRODUCTION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
The present study was performed to investigate an iron oxide–based blood pool contrast medium in terms of its potential for MR angiography in the coronary arteries. Unlike all other iron oxide–based contrast media developed to date, which are coated with a polymer (eg, dextran, carboxydextran, polyethylene glycol), VSOP-C91 (Ferropharm, Teltow, Germany), very small superparamagnetic iron oxide (SPIO) particles, is coated with a monomer (citrate). The particles are characterized by a favorable ratio of T2 relaxivity (r2) to T1 relaxivity (r1) in combination with a long blood pool half-life (6,7), which make it a good candidate for blood pool imaging. The purpose of our study was to evaluate the signal-enhancing characteristics of the very small SPIO particles as a blood pool contrast medium for MR angiography in the coronary arteries in comparison with two low–molecular-weight contrast media: an unspecific agent with extracellular distribution (gadopentetate dimeglumine, Magnevist [0.3 mmol of gadolinium per kilogram of body weight]; Schering, Berlin, Germany) and an agent with a low protein-binding capacity, which is considered beneficial for contrast material–enhanced MR angiography (gadobenate dimeglumine, Multihance [0.2 mmol Gd/kg]; Bracco-Byk Gulden, Konstanz, Germany) (8).
| MATERIALS AND METHODS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Before MR imaging, anesthesia was induced by means of intramuscular injection of 15 mg/kg of ketamine hydrochloride (Ketamin [500 mg]; Curamed Pharma, Karlsruhe, Germany), 0.2 mg/kg droperidol (Dehydrobenzperidol; Janssen-Cilag, Neuss, Germany), and 0.2 mg/kg midazolam hydrochloride (Dormicum; Hoffmann-La Roche, Grenzach-Whylen, Germany). After the anesthetics took effect, a 6.0–6.5-mm endotracheal tube (Mallinckrodt Laboratories, Athlon, Ireland) was inserted into the trachea to maintain anesthesia with a mixture of 2%–3% isoflurane (Forene; Abbott, Wiesbaden, Germany) and medical oxygen. Anesthesia was maintained by using an electronic system for controlling ventilation and anesthesia (ADS 1000; Engler Engineering, Hialeah, Fla). The contrast material was administered via an indwelling 22-gauge cannula (Optiva 2; Ethicon, Pomezia, Italy), which was previously inserted into a marginal ear vein.
Contrast Medium and Doses
The very small SPIO particles were prepared as described elsewhere (6). The particles have a core diameter of 5 nm, as determined with transmission electron microscopy, and a hydrodynamic diameter of 7 nm, as determined with laser light scattering (Zetasizer 3000; Malvern Instruments, Malvern, Worcestershire, England). The relaxivities determined at three field strengths are given in Table 1. At 0.47 and 0.94 T (Minispec PC 100 and Minispec MQ 40, respectively; Bruker, Karlsruhe, Germany), the relaxivities were determined with the standard measuring and analysis methods of the equipment. At 1.5 T (Magnetom Vision; Siemens Medical Systems, Erlangen, Germany), r1 was determined with a turbo inversion-recovery sequence, and r2 was determined with a spin-echo multiecho sequence in combination with a non–section-selective 180° refocusing radio-frequency pulse. VSOP-C91 designates the 91st variant in the optimization process in developing this contrast medium. The core size and the citrate coating were systemically varied to achieve a minimum ratio of r2 to r1, as a physical parameter, while at the same time maximizing blood half-life, as a biologic parameter. In addition, it was necessary to pay attention to tolerability of the agent. A preliminary study of the tolerance of the very small SPIO particles in rats (n = 5) found neither an impairment of the animals’ well being nor any macroscopic or microscopic pathologic organ changes after bolus injection of doses as large as 5 mmol/kg. The pigs received the very small SPIO particles at a dose of 60 µmol Fe/kg. This dose was found to produce adequate intravascular signal enhancement in a prior dose optimization study in rats (7). The very small SPIO particles have a blood half-life of 71 minutes ± 9 (SD) at 60 µmol Fe/kg (determined in five rats) and a ratio of r2 to r1 of 2.1 L/mmol · sec-1 (determined at 1.5 T).
|
Five examinations were performed with each contrast medium. MR imaging was performed after intravenous bolus injection of the contrast medium at a flow rate of 1 mL/sec followed by a 20-mL 0.9% saline flush at the same rate.
MR Imaging Protocols
All MR imaging examinations were performed with a 1.5-T MR imager, with a maximum gradient amplitude of 25 mT/m, and the commercially available four-element body phased-array coil. The hearts of the pigs were imaged with a three-dimensional fast low-angle shot, or FLASH, sequence optimized for a minimum acquisition time with use of a small field of view. This was achieved by using echo-planar imaging gradients with a minimum rise time of 300 µsec. The zero-filling technique was used with the following parameters: repetition time msec/echo time msec of 4.5/1.7; flip angle, 25°; receiver bandwidth, 325 Hz per pixel; transverse orientation; rectangular field of view, 160 x 320; matrix, 140 x 512 (112 measured phase-encoding steps, 140 reconstructed phase-encoding steps); slab thickness, 40 mm; 20 partitions acquired; 40 partitions reconstructed; effective section thickness, 1 mm. The pulse sequence was programmed to acquire the 20 partitions during every pulse cycle during a 90-msec interval, with the number of cardiac cycles required equal to the number of phase-encoding steps (ie, 112).
A presaturation technique was not used, to ensure comparability of the three contrast media on the basis of quantitative analysis, including myocardial enhancement. The pulse signal was registered with a pulse oximeter at the lower hind limb, and the trigger delay was set at 0 msec. The heart rates of the animals during MR imaging ranged between 70 and 96 min-1 (mean, 83 min-1 ± 12). Comparison with the electrocardiogram showed that data acquisition occurred during end diastole, approximately 300–400 msec after the R wave. Total acquisition time was approximately 60–90 seconds, depending on the heart rate of the pigs. All images were acquired during expiratory standstill, which was induced by manually switching off the artificial ventilation system. Images were acquired immediately before the contrast material was administered and at 5- to 10-minute intervals thereafter for as long as 35 minutes. The first postcontrast acquisition was started with the injection.
Analysis
Quantitative analysis.—Signal intensities of the blood in the ventricle, the left coronary artery (proximal portion), and the left ventricular myocardium were determined by means of standard region-of-interest measurements by two readers (C.A., D.K.) on the corresponding MR angiograms from near the base of the heart for all time points. Taking into account adjoining images, the two readers together determined the position and size of each region of interest (with round to oval configuration), to exclude any influence of partial volume effects and motion artifacts on the measurements. The selected images also contained a region of interest placed outside the body, to measure the signal intensity of noise, that contained no motion artifacts in the phase-encoding direction. The sizes of the regions of interest placed in the left ventricular cavity, left coronary artery, left ventricular myocardium, and outside the body (noise) were 125–455, 2.8–5.6, 17–27, and 1,920–3,560 mm2, respectively. At all time points, signal-to-noise ratios (SNRs) were calculated for the signal intensity (SI) of blood (ventricular cavity, left coronary artery) and myocardium, and contrast-to-noise ratios (CNRs) were calculated for the contrast between blood (B) and myocardium (M): CNR = (SIB - SIM)/SIN, where N is noise.
Qualitative analysis.—For qualitative analysis, the second postcontrast acquisition of each examination (n = 15) was assessed by two experienced readers (reader 1, S.W.; reader 2, M.T.), who were blinded to the type of contrast medium administered. The following criteria were assessed: (a) Overall image quality: 1, pronounced motion artifacts, pronounced contour blurring; 2, moderate motion artifacts, moderate contour blurring; 3, minimum motion artifacts, sharply delineated contours; 4, no motion artifacts, very sharply delineated contours. (b) Contrast between myocardium and coronary artery lumen or ventricular lumen: 1, slight; 2, moderate; 3, good; 4, excellent. (c) Continuity of the coronary arteries, separately for the right and left coronary artery: 1, poor or fragmentary visualization of the coronary artery; 2, depiction of the proximal two-thirds of the coronary artery with no more than one disruption; 3, the entire course of the coronary artery seen without disruption; 4, excellent depiction of the entire course of the coronary artery to the apex of the heart. (d) Visualization of lateral branches of the coronary arteries: 1, no lateral branches visible; 2, origins of individual lateral branches just barely visible; 3, longer segments as large as two lateral branches visible; 4, longer segments of at least three lateral branches visible. The scores were added separately for the two readers. Qualitative analysis was performed with a workstation (Magic View; Siemens Medical Systems), and the images were evaluated together by the two readers.
Statistical analysis.—A nonparametric analysis of variance was performed to take into account the correlated nature of the measurements (9,10). For quantitative analysis, a two-factorial analysis of variance of the factors contrast medium and time was performed for each tissue. Owing to the limited number of samples, only three postcontrast time points could be taken into account. The 0-minute (immediately after injection), 5-minute, and 35-minute time points were chosen for data analysis. The Wilcoxon test for paired samples was used to identify any changes in heart rates of the pigs during the examinations or between the groups of animals examined with the three contrast media. For qualitative analysis, a two-factorial analysis of variance of the factors contrast agent and reader was performed for each tissue. Differences with a P value less than .05 were considered significant.
| RESULTS |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
Quantitative Analysis
The quantitative results, expressed as SNRs, are presented in Table 2. The increase in SNRs in the lumen of the coronary artery after administration of gadopentetate dimeglumine, gadobenate dimeglumine, and the very small SPIO particles peaked at the first postcontrast acquisition (94%, 103%, and 102%, respectively). Differences among the contrast media at the postcontrast time points were not significant. In the myocardium, the highest increase in SNRs occurred at the first postcontrast acquisition: 83% for gadopentetate dimeglumine, 83% for gadobenate dimeglumine, and 29% for the very small SPIO particles. SNRs with gadobenate dimeglumine were significantly higher than those with the very small SPIO particles at all time points, but differences were not significant between the two gadolinium-based compounds. Differences between precontrast and all postcontrast SNRs in the myocardium and ventricular volume were significant for all contrast media.
|
|
|
|
|
|
|
|
|
| DISCUSSION |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|
All MR angiographic contrast agents currently available for clinical use are low–molecular-weight compounds that are rapidly cleared from the blood and distributed in the extracellular space after intravenous administration. To assess the coronary arteries, first-pass imaging of the heart is necessary with an electrocardiography-triggered sequence performed during a 20–30-second breath hold (3–5). Disadvantages of these techniques include that they require exact positioning of the three-dimensional volumes beforeacquisition and that their spatial resolution is low owing to the limited time available. To completely visualize the heart with high spatial resolution, the time of data acquisition surpasses both the first pass of the contrast medium and the breath-hold capacity of the patient. To overcome these limitations of extracellular low–molecular-weight contrast media, Zheng and coworkers (11) suggest use of protocols with slow infusion of the contrast agent to achieve adequate enhancement of the intravascular signal during the entire imaging time with a navigator-gated three-dimensional acquisition during breathing. They conclude that constant T1 shortening during the entire acquisition period is required to ensure good visualization of the coronary arteries with good boundary definition. They imply that a blood pool agent is likely to yield good results, particularly in prolonged imaging, as with navigator gating.
In the present study, we investigated use of a compound from a group of ultrasmall SPIO particles as a blood pool contrast medium for coronary MR angiography (7). The very small SPIO particles are coated with a monomer (6) rather than polymers, as are used in all other ultrasmall SPIO particles to date. Monomer coating results in a particle structure that is easy to define in physical and chemical terms. The overall particle diameter of the very small SPIO particles can be varied between 2 and 10 nm, and r1 and r2 can be varied over a wide range in a reproducible manner.
VSOP-C91 is coated with citrate as the monomer. Citrate is an endogenous substance that occurs in mammals and can be metabolized by them, which provides optimal conditions for excellent tolerance of the compound. All other iron oxide–based blood pool contrast media currently undergoing preclinical or clinical testing are coated with polymeric material. Dextran coating limits the tolerance of the compound, because it can be administered only as a slow infusion (12). In general, all polymers, even the well-tolerated polyethylene glycol, can produce adverse reactions (13–15). As with conventional polymer-coated iron oxide particles (16), the citrate-coated agent is primarily cleared by the mononuclear phagocytosing system after intravenous injection. Findings in preliminary studies suggest that phagocytosis by Kupffer cells in the liver and macrophages in the spleen produces a pronounced transient signal loss in the liver and spleen. Doses as large as 75 µmol Fe/kg also result in some transient accumulation of the very small SPIO particles in other organs, including the myocardium, but this has no effect on MR imaging.
The very small SPIO particles have an overall particle diameter of about 7 nm. Such a small particle is an important prerequisite for prolonged intravascular retention (17). In addition, the compound exhibited favorable relaxivities, which are crucial for the signal-changing properties of a contrast medium. Compared with gadolinium-based blood pool media, ultrasmall SPIO particles have a similar r1 but markedly higher r2. This property may already reduce the signal at the doses that are required to effectively shorten r1, which is especially deleterious when use of extremely short echo times is not possible (18). It is important that ultrasmall SPIO particles have a high r1 and a low ratio of r2 to r1. At 0.47 T, the very small SPIO particles have an r1 of 19 L/mmol · sec-1 and an r2 of 29 L/mmol · sec-1 with a resultant r2/r1 ratio of 1.5. This ratio is more favorable than that of other ultrasmall SPIO particles. The r2/r1 ratios reported for other contrast media at 0.47 T are 1.75 for NC100150 (Clariscan; Nycomed-Amersham, Oslo, Norway), 2.5 for AMI-227 (Combidex; Advanced Magnetics, Cambridge, Mass), and 2.3 for SH U 555C (Resovist-S; Schering) (19–21). The only other contrast agent with an r2/r1 ratio as favorable as that of the very small SPIO particles is a gadolinium-containing macromolecular contrast agent (SH L 643A, Gadomer-17; Schering) (r2/r1 ratio, 1.55) (22). It is well established that the relaxivities of contrast media may vary considerably with the magnetic field strength. Depending on their composition, iron oxide–based contrast media show a more or less pronounced decrease in r1 and increase in r2 at increasing field strengths (23).
The preliminary investigations of the very small SPIO particles performed at three field strengths revealed only minimum changes in r1 and r2 (Table 1). Few data are available about the dependence of relaxivity on field strength for the other compounds currently undergoing testing as blood pool contrast media. At 2.0 T, the ultrasmall SPIO particles SH U 555C is reported to have an r1 of 7.8 L/mmol · sec-1 and an r2 of 87 L/mmol · sec-1 (r2/r1 = 11.1), whereas SH L 643A has an r1 of 14.7 L/mmol · sec-1 and an r2 of 21.4 L/mmol · sec-1 (r2/r1 = 1.46) (22). To the extent that relaxivities reported in the literature, which may have been measured with different techniques, can be compared with those of the very small SPIO particles, the latter are closer to those of a macromolecular gadolinium-containing blood pool contrast medium (SH L 643A) than to those of other ultrasmall SPIO particles blood pool media.
Our study involved a direct comparison of a blood pool contrast medium with low–molecular-weight gadolinium-based compounds for coronary MR angiography both early after injection and during the equilibrium phase. The results obtained by using a sequence without preparation pulses show that although the very small SPIO particles produce an intravascular signal enhancement comparable to that of the low–molecular-weight compounds gadopentetate dimeglumine and gadobenate dimeglumine in the early phase and during equilibrium, the contrast between the coronary arteries and the myocardium is improved as a result of the significantly lower myocardial signal increase. This enables good qualitative assessment of depiction of the coronary arteries and its branches. However, the drawback of low–molecular-weight contrast media that results from the undesired myocardial signal increase due to their rapid extravasation can be compensated, at least to some extent, by using various presaturation techniques. Zheng and coworkers used an inversion pulse prepared three-dimensional gradient-echo sequence to suppress the postcontrast signal of nonblood tissue (11).
Another option for suppressing the background signal is to apply a series of prepulses that have the same excitation angle as that used for data acquisition. This so-called dummy scanning generates an equilibrium magnetization with suppression of background signal, as occurs throughout data acquisition in non–electrocardiography-triggered contrast-enhanced MR angiography in other body regions (24). We deliberately chose not to use such prepulses so we could measure the signal enhancement of the myocardium after injection of the different contrast media with a high degree of sensitivity. There is no doubt, however, that the application of such prepulses to suppress myocardial signal, although it is already low, can also further improve the quality of coronary MR angiography with the very small SPIO particles. On the other hand, it must be noted that the kinds of prepulses used for suppressing myocardial background signal reduce the SNR. Coronary MR angiography would therefore benefit from a contrast medium that can be used without prepulses.
Other blood pool contrast media presumably have similar advantages over low–molecular-weight agents as those found for the very small SPIO particles in the present study. It would therefore be of interest to directly compare blood pool contrast media by using the standardized model of the pig coronary arteries to identify the most suitable candidate. To our knowledge, such a study has not yet been published. A blood pool contrast medium, such as VSOP-C91, would be advantageous not only in coronary MR angiography with breath-hold technique, but would also allow imaging of the entire coronary tree with repeat acquisitions. The prolonged contrast between the coronary artery and the myocardium could also be exploited for navigator-controlled imaging techniques with acquisition times as long as 20 minutes, which would improve the image quality in patients with a reduced breath-hold capacity.
In summary, VSOP-C91, an iron oxide–based contrast medium coated with the monomer citrate, has favorable properties as a blood pool contrast medium for MR angiography in the coronary arteries that result from its long intravascular retention in combination with its favorable r2/r1 ratio. The effective intravascular signal enhancement with almost no myocardial enhancement that resulted from these properties allowed excellent depiction of the coronary arteries in this experimental study in pigs.
Practical application: Our experimental results show that VSOP-C91 is a suitable contrast medium for performing MR angiography in the coronary arteries. In a clinical setting in the future, this agent might have a role in improving noninvasive diagnostic assessment of the coronary arteries. The prolonged and selective intravascular signal enhancement induced by the substance could be used for imaging in the entire coronary tree during repeated breath holds or in prolonged volume-covering navigator-gated image acquisition during free breathing.
| ACKNOWLEDGMENTS |
|---|
| REFERENCES |
|---|
| Top ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES |
|---|