| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Department of Radiology, University of California San Francisco, 505 Parnassus Ave, L308, San Francisco, CA 94143-0628. From the 2001 RSNA scientific assembly. Received September 11, 2001; revision requested November 1; final revision received March 6, 2002; accepted March 28. Address correspondence to M.S. (e-mail: maythem.saeed@radiology.ucsf.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Twenty rats underwent 45 minutes of coronary artery occlusion and 3 hours of reperfusion. Sequential T1-weighted spin-echo MR images were acquired in 10 rats to assess leakage of Clariscan into myocardium over time. Ten other rats underwent the same duration of occlusion and reperfusion (3 hours) so that the extent of microvascular injury in the entire heart could be measured and correlated with infarction and area at risk at necropsy. The Student t test and Bland-Altman method were used for data analysis.
RESULTS: Clariscan improved visualization of regions with transmural and nontransmural microvascular injury. Accumulation of Clariscan was best reflected by the mean ratios of signal intensity in injured myocardium to that in normal myocardium measured before (0.98 ± 0.01 [standard error of the mean]) and after (1.34 ± 0.04) injection. At 15 minutes after injection, the size of the enhanced region remained constant over the course of observation. The mean size of the hyperenhanced region (44% of the left ventricle ± 2) was significantly (P < .001) larger than the mean size of true infarction at necropsy (29% ± 3) but smaller than the mean size of the area at risk (50% ± 2).
CONCLUSION: Clariscan has potential for estimating the spatial extent of microvascular injury in ischemically injured myocardium and may be useful as a marker of microvascular injury after thrombolytic therapy.
© RSNA, 2002
Index terms: Animals • Iron • Magnetic resonance (MR), contrast media • Magnetic resonance (MR), experimental studies • Myocardium, infarction, 511.814
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Contrast material-enhanced magnetic resonance (MR) imaging has recently been used in patients for (a) discriminating transmural from nontransmural infarction (5), (b) delineating occlusive and reperfused infarctions (6), and (c) assessing myocardial viability (5,7). All of these applications have incorporated the use of standard extracellular (low-molecular-size) MR contrast media (ie, gadolinium chelates). However, extracellular MR contrast agents rapidly distribute into the extravascular space, and this distribution eliminates their potential use in assessing microvascular integrity.
For assessment of microvascular injury, blood pool MR contrast media that allow steady-state measurements of microvascular integrity are desirable (8). In biodistribution studies, blood pool MR contrast media have been shown to remain largely confined to the intravascular space of normal myocardium (8) and to slowly accumulate in regions with ischemically injured microvessels (9). To demonstrate temporal changes in the leakage of microvessels, steady-state measurements are needed. Clariscan (feruglose NC100150 injection, Nycomed-Amersham, Oslo, Norway) is a blood pool agent that is being evaluated in clinical trials in the United States and Europe. Clinical studies have shown that Clariscan is highly tolerated by humans, and this agent has been successfully used in coronary MR angiography and the assessment of myocardial perfusion (10,11). Furthermore, the diagnostic potential of this agent for characterization of microvascular integrity in different types of tumors has been demonstrated (12,13).
To our knowledge, Clariscan has not been previously tested for its potential in characterizing and estimating the spatial extent of microvascular injury in reperfused infarction. The purposes of the current study were (a) to assess the value of Clariscan for characterizing and estimating the spatial extent of microvascular injury in reperfused ischemically injured rat myocardium and (b) to compare the size and transmural extent of Clariscan-enhanced regions seen at MR imaging with the size and transmural extent of myocardial infarction and areas at risk seen at postmortem histochemical staining.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
(ie, the partition coefficient).
Experimental Protocol
Care and treatment of experimental animals were performed in strict accordance with National Institutes of Heath guidelines. The experimental protocol was approved by the committee on animal research at this institution. Twenty rats (Simonsen Laboratories, Gilroy, Calif) that ranged in body weight from 295 to 320 g and had a mean body weight of 310 g ± 5 (standard error of the mean) were anesthetized with intraperitoneal injection of 50 mg of sodium pentobarbital (Nembutal Sodium Solution; Abbott Laboratories, North Chicago, Ill) per kilogram of body weight.
After tracheotomy was performed in each animal, the animals were mechanically ventilated. Thoracotomy of the left side of the thorax was performed in each animal, and the left anterior coronary artery was occluded after a snare loop was placed around the vessel. Animals were subjected to 45 minutes of coronary artery occlusion and 3 hours of reperfusion. Baseline MR images were acquired 3 hours after reperfusion.
For characterization of microvascular injury and sizing of the spatial extent of microvascular injury, two groups of 10 rats each were studied. Group 1 rats received 0.05 mmol of Clariscan per kilogram of body weight (ie, 3 mg of iron per kilogram) after baseline T1-weighted spin-echo MR images had been acquired. The contrast agent was manually injected for 30 seconds through an intravenous tail vein catheter. In group 1 rats, signal intensity–over-time curves at a single section location at the midventricular level were used to determine the optimal time for measuring the extent of microvascular injury. A previous study in pigs has shown that the concentration of Clariscan in plasma decreases the T1 of blood monoexponentially in a dose-dependent fashion (15). The plasma half-life of Clariscan is 45–100 minutes. The imaging protocol used in this study was performed in less time than the duration of the plasma half-life of this particular contrast agent.
In normal myocardium, blood pool agents occupy the blood pool and reach a state of equilibrium in the blood within 3 minutes. However, in myocardium that contains injured microvessels, blood pool agents distribute by diffusion and convection in the extravascular space. The time for these agents to reach equilibrium distribution in injured myocardium depends on the size of the infarcted region and the patency of the microvessels (8,9).
In this study, the accumulation of contrast medium in the ischemically injured region was monitored (G.A.K., M.S.) for 45 minutes after injection with serial MR images obtained at the midventricular level. The location of the midventricular section was determined by measuring the long-axis view of the heart (1.1–1.3 cm).
In group 2, 10 animals were subjected to the same duration of occlusion (45 minutes) and reperfusion (3 hours) before MR imaging. An identical dose of 3 mg of iron per kilogram of body weight of Clariscan was administered. At the optimal time (ie, 25 minutes after injection), which had been determined from data acquired in group 1, group 2 rats were assessed for the extent of microvascular integrity in the entire heart. A series of short-axis images at contiguous section locations were acquired to enable measurement of the extent of microvascular injury from the apex of the heart to the base of the left ventricle. The circumferential and transmural extents of microvascular injuries on contrast-enhanced MR images were compared with infarction size seen with triphenyltetrazolium chloride (TTC) staining of tissue and with the area at risk of infarction if there was no reperfusion.
The sizes of the enhancing regions in transmural and nontransmural injuries were compared (G.A.K., M.S.) with areas at risk and with areas of true infarction measured after sacrifice of the animal (16). Values are given as percentages of left ventricular surface rather than in milligrams to avoid the confounding effect of differences in rat body weights and heart weights.
MR Imaging
Electrocardiographically gated MR images were obtained with a 2-T system (Omega CSI; Bruker Instruments, Fremont, Calif). T1-weighted spin-echo MR images were acquired in all 20 rats before and serially after injection of the contrast medium (G.A.K., M.S., M.F.W.). Acquisition parameters were as follows: repetition time msec/echo time msec, 300/12; matrix size, 256 x 128 (interpolated to 256 x 256 during Fourier transformation); field of view, 50 x 50 mm; section thickness, 2 mm; number of acquisitions, four; and scan time, 2.5 minutes.
Signal intensity was measured before and at 1, 5, 15, 20, 25, and 45 minutes after injection. Signal intensity measurements were obtained in remote normal myocardium, the septum, ischemically injured myocardium, and the skeletal muscles of the back (ie, transversospinal and trapezius muscles). In group 2 animals, images were acquired 25 minutes after injection to cover the entire heart. Two reviewers (G.A.K., M.S.) interpreted MR images in consensus. The signal intensity behavior of left ventricular chamber blood adjacent to the injured myocardium was observed before and after the injection of the contrast medium. The homogeneity of the enhancing regions was also determined.
Postmortem Measurements
After imaging, the coronary artery was reoccluded in all rats, and 0.2 mL of phthalocyanine blue dye (Engelhard, Louisville, Ky) was injected intravenously to demarcate the area at risk. The artery was reoccluded to define the territory of ischemia by preventing the entrance of the blue dye. The left ventricle of each rat was transversely cut into 2-mm-thick slices, corresponding to the section thickness of the MR images; the slices were then scanned by using a flatbed scanner (Silverscanner IV; LaCie, Hillsboro, Ore). The slices were incubated in 2% TTC for 10 minutes to define the infarcted region. The size of the area at risk and the size of infarction were quantified by two readers (G.A.K., M.S.) in consensus by using a public-domain image-analysis software (NIH Image, developed at the National Institutes of Health and available on the Internet at cmex-www.arc.nasa.gov/CMEX/ data/Tutorial/DocInfo.htm).
Statistical Analysis
Signal intensity–over time curves were calculated for normal and injured myocardium, as well as for skeletal muscle. All data are presented as means ± standard errors of the mean except those yielded by Bland-Altman analysis. The statistical differences among the sizes of regions of enhancement on T1-weighted images and the sizes of areas at risk and areas of infarction at histochemical staining were determined by using the Student t test. If the analysis showed an overall P value less than .05, the Scheffé F test was performed as a post-hoc test. Linear regression analysis and Bland-Altman analysis (results of which are presented as means ± 2 SDs) were performed to determine the correlation coefficient and agreement between MR imaging and histomorphometry. The null hypothesis was rejected when P was less than .05.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
The conspicuity of ischemically injured myocardium was improved after Clariscan was administered, as reflected by an increase in the mean signal intensity ratio. (The mean ratio of signal intensity observed in ischemically injured myocardium to that observed in normal myocardium was 0.98 ± 0.01 before and 1.34 ± 0.04 after injection of Clariscan.) The signal intensity ratio did not change significantly (P = .40) in the last 30 minutes of the observation period (mean signal intensity ratio, 1.45 ± 0.08 at 15 minutes and 1.42 ± 0.09 at 45 minutes). Furthermore, the apparent mean size of a hyperenhanced region in a single section did not change between 15 and 45 minutes after injection (38% of left ventricle ± 3 at 15 minutes vs 38% ± 2 at 45 minutes, P = .40).
The leakage of Clariscan in the injured region was essentially uniform except in two animals (20%), in which contrast medium was leaked in a heterogeneous manner, which resulted in the appearance of a dark core surrounded by a high-signal-intensity zone in the first 5 minutes. It has previously been shown that this central dark zone does not fluoresce after administration of a thioflavin S fluorescent tracer; this constitutes the no-reflow phenomenon (18). Furthermore, in four (40%) of the animals, nontransmural enhancement was observed on contrast-enhanced MR images, indicating that the microvascular injury was limited to submyocardial and middle myocardial areas (Fig 3). Figure 4 shows the correlation between transmural and nontransmural microvascular injury as observed on Clariscan-enhanced T1-weighted spin-echo MR images, the areas found to be at risk after phthalocyanine blue dye infusion, and the areas of true infarction seen with TTC staining.
|
|
|
|
Results of Bland-Altman analysis revealed that Clariscan-enhanced regions were overestimated by 15% ± 7.0 (mean ± 2 SDs) at MR imaging compared with areas of true infarction found at TTC staining, but were underestimated by -5% ± 9 compared with areas at risk delineated by phthalocyanine blue dye administration (Fig 7). Good correlation was found between Clariscan-enhanced regions and areas of true infarction observed at TTC staining (r = 0.89; P < .006; Y = 16 + 0.976 · X, where Y indicates the value on the y axis of the linear regression analysis test and X indicates the value on the x axis of the test) and between Clariscan-enhanced regions and areas at risk delineated by phthalocyanine blue dye administration (r = 0.73, P < .016, Y = -9.8 + 1.08 · X) (Fig 8). The enhanced region was significantly (P < .01) smaller (44% ± 2) than the true area at risk (50% ± 3); this may be attributable to subregional microvascular hyperpermeability (Fig 5).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Assessment of Microvascular Integrity with MR Imaging
Until recently, an in situ technique for making an absolute distinction between functional and nonfunctional microvessels has not been available. The advent of blood pool agents has opened up important avenues toward the use of more specific agents in diagnostic imaging. These agents are characterized by a biodistribution that is limited to the vascular space. This property is potentially useful for several diagnostic applications, such as MR angiography and MR depiction of microvascular permeability disorders in myocardium and tumors.
In the current study, microvascular injury in ischemically injured myocardium was characterized by increasing signal intensity after administration of contrast material. This pattern of signal intensity increase in ischemically injured myocardium suggests that (a) Clariscan was delivered into the reperfused ischemically injured region, (b) Clariscan leaked into extravascular space via injured microvessels, and (c) the distribution volume of Clariscan was greater in the injured region than in normal myocardium.
Assessment of myocardial perfusion and microvessel obstruction has previously been studied with rapidly diffusible extracellular MR contrast media in conjunction with dynamic MR imaging (19–22). However, because of their fast-diffusing nature, these agents cannot be used for detection of microvascular hyperpermeability or loss of vascular integrity in ischemically injured myocardium. Furthermore, the sizes of enhanced regions vary after the injection of extracellular MR contrast media due to the rapid washout of the agents. Ni et al (23) and Oshinski et al (24) found in dogs and rats, respectively, that the size of regions enhancing at gadolinium-enhanced MR imaging rapidly declines by 20%–40% in the first 20–30 minutes. A number of interrelated factors are responsible for the variability in the extent of the gadolinium-enhanced region, including regional blood flow, size of the injured region, interstitial volume in the periinfarction zone, and severity of myocardial injury.
Estimation of the Spatial Extent of Microvascular Injury
In the current study, we found that, unlike results obtained with extracellular MR contrast agents, the magnitude and extent of Clariscan-enhanced regions in transmural and nontransmural microvascular injuries remained constant for 45 minutes of observation. One possible reason for the constancy of the size of the enhanced region may have been that the contrast medium escaped the blood pool and resided in the interstitium. Slow clearance of Clariscan from the blood pool and the large size of its particles, which hampers reentry of the agent into the capillaries, can explain the constant size of the enhanced region.
One of the major findings of the current study is that the area of enhancement seen after administration of the blood pool agent corresponds to an overestimation of the size of true infarction but an underestimation of the size of the area at risk. The average overestimation of infarcted myocardium with Clariscan-enhanced MR imaging was approximately 15%, which is almost identical to the overestimation observed after administration of the extracellular MR contrast medium gadopentetate dimeglumine (12%–20%) in rats, cats, and dogs (16,23–28).
Clariscan also has the potential to enable discrimination between transmural and nontransmural microvascular injury. The overestimation of nontransmural microvascular injury (17%) at Clariscan-enhanced MR imaging was significantly greater than the overestimation of transmural injury (12%, P < .01). The reason for the difference is not clear at this stage. In the current study, results of TTC staining were used as the standard of diagnosis for delineating myocardial infarction (29). The region in which infarction was overestimated at contrast-enhanced MR imaging (calculated by "subtracting" the region of infarction observed with TTC staining from the region that enhanced at MR imaging) most probably represents the periinfarction zone described in previous reports (16,28).
Because of the differential enhancement of the periinfarction zone compared with the enhancement of remote normal myocardium, it has been suggested that the capillaries in this region are hyperpermeable. The apparent leakage of Clariscan from microvessels in the viable periinfarcted zone is in agreement with results of physiologic (30) and histologic (31) studies.
The leakage of macromolecules from microvessels observed in this experimental model of reperfused infarction in rats has previously been described with the experimental blood pool MR contrast medium albumin-(biotin)10-gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA)-25 (9). Use of this experimental blood pool agent enabled microscopic confirmation of the existence of microvascular permeability to macromolecules, as well as demonstration of the no-reflow phenomenon. Use of albumin-(biotin)10-Gd-DTPA-25 removed all ambiguity between the distinction of MR contrast medium and distribution of the biotin complex used for histopathologic staining (9).
To our knowledge, this is the first study of MR imaging that demonstrates that microvessels in the periinfarction zone have been impaired and have become hyperpermeable. This finding may explain the overestimation of infarction seen when T2-weighted or extracellular contrast material–enhanced spin-echo MR images are used (16,23,32). Results of recent studies with dobutamine thallium 201 single photon emission computed tomographic imaging and cine MR imaging support the notion that the enhanced periinfarction zone is viable and functional (25,33) and benefits from cardioprotective therapy (34).
Furthermore, the results of the present study suggest that microvascular injury precedes myocardial necrosis. This finding is in agreement with those of recent studies in which invasive techniques were used (31,35). One functional study of microvascular permeability indicated that myocardial ischemia as short as 15–20 minutes in duration causes a significant increase in protein extravasation in the reperfused ischemic myocardium (35). Results of another recent study in rats indicated that apoptosis is first seen in the endothelial cells of small coronary vessels (31). These investigators demonstrated that the radial spread of apoptosis to surrounding myocytes suggests that reperfusion induces the release of soluble proapoptotic mediators from endothelial cells that promote apoptosis in myocytes (31).
Blood Pool MR Contrast Media
The degree to which blood pool contrast agents cross the normal capillary wall is limited and depends on regional permeability of microvessels (eg, ischemic injury, inflammation, and tumor) and the physicochemical characteristics (eg, size, charge, and molecular shape) of the blood pool agents themselves (36). Once blood pool agents permeate the microvessels, they gain access to a large water pool (interstitial and cellular) that contributes to differential enhancement.
The diagnostic potential of several blood pool contrast agents at MR imaging has been tested experimentally (8–15). However, the first agents tested (eg, Gd-DTPA–albumin, Gd-DTPA–polylysine, Gd-DTPA–dextran) are unlikely to be further evaluated in clinical trials because of their incomplete elimination and potential to be toxic or provoke an immunologic response. More recently, highly tolerable MR contrast agents such as Clariscan and magnesium 325 have been developed and have entered clinical trials (9,10,37). Clariscan has certain advantages over the other agents. Foremost is the fact that the clearance of Clariscan favors clinical utility. This agent remains in the blood pool for more than 90 minutes, resulting in prolonged enhancement of myocardium (9,10). This factor may explain the potential of this agent in delineating occlusive infarction (38), a potential that has not been observed with other superparamagnetic iron particles such as americium 227 (39).
The results of this study indicate that Clariscan, a blood pool agent currently in phase III clinical trials, is useful for estimating the spatial extent of microvascular injury in ischemically injured myocardium when a steady-state MR contrast medium approach is used. Clariscan has the potential to demonstrate transmural and nontransmural microvascular injury in reperfused infarctions. Contrast-enhanced MR imaging may be useful as a marker of reperfusion and angiogenesis after thrombolytic and gene therapy.
Limitations of the Study
The major limitations of the current study were as follows: (a) We did not use kinetic modeling (13,14) to measure the leakage of Clariscan in the infarcted and periinfarcted regions. (b) A shorter period of ischemia (without infarction) must be induced and observed to document whether loss of microvascular integrity precedes myocardial infarction. However, the apparent overestimation with contrast-enhanced MR imaging observed in the present study is in agreement with results of physiologic and histologic studies by Dauber et al (30) and Scarabelli et al (31). (c) First-pass measurements were not performed in the current study to define the ischemic region.
Practical application: An advantage of the steady-state MR contrast medium approach over the first-pass imaging approach is that MR images can be obtained in a wide time window after the administration of contrast material. Furthermore, the ability to discriminate transmural from nontransmural microvascular injuries could prove to have clinical utility for selecting patients eligible for therapeutic interventions more appropriately, for monitoring treatment, and for targeting locally delivered therapies.
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, M.S., G.A.K., C.B.H.; study concepts and design, M.S.; literature research, M.S., G.A.K.; experimental studies, M.S., G.A.K., M.F.W.; data acquisition, G.A.K., M.S., M.F.W.; data analysis/interpretation, G.A.K., M.S.; statistical analysis, G.A.K., M.S.; manuscript preparation, G.A.K., M.S.; manuscript definition of intellectual content, editing, and revision/review, M.S., G.A.K., C.B.H.; manuscript final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  D. E. Sosnovik, M. Nahrendorf, and R. Weissleder Molecular Magnetic Resonance Imaging in Cardiovascular Medicine Circulation, April 17, 2007; 115(15): 2076 - 2086. [Full Text] [PDF]
|
|
|
|
 |
|
 M. Saeed, O. Weber, R. Lee, L. Do, A. Martin, D. Saloner, P. Ursell, P. Robert, C. Corot, and C. B. Higgins Discrimination of Myocardial Acute and Chronic (Scar) Infarctions on Delayed Contrast Enhanced Magnetic Resonance Imaging With Intravascular Magnetic Resonance Contrast Media J. Am. Coll. Cardiol., November 21, 2006; 48(10): 1961 - 1968. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Jackowski, A. Christe, M. Sonnenschein, E. Aghayev, and M. J. Thali Postmortem unenhanced magnetic resonance imaging of myocardial infarction in correlation to histological infarction age characterization Eur. Heart J., October 2, 2006; 27(20): 2459 - 2467. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. A. Krombach, C. B. Higgins, M. Chujo, and M. Saeed Gadomer-enhanced MR Imaging in the Detection of Microvascular Obstruction: Alleviation with Nicorandil Therapy Radiology, August 1, 2005; 236(2): 510 - 518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. R. Edelman Contrast-enhanced MR Imaging of the Heart: Overview of the Literature Radiology, September 1, 2004; 232(3): 653 - 668. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Fujiwara, K. Togashi, T. Yamaoka, A. Nakai, A. Kido, S. Nishio, T. Yamamoto, H. Kitagaki, and S. Fujii Kinematics of the Uterus: Cine Mode MR Imaging RadioGraphics, January 1, 2004; 24(1): e19 - e19. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
= P < .001 for difference between extent of Clariscan-enhanced regions and extent of areas of true infarction.
