Info

'Watanabe heriditable hyperlipidemic rabbits ^New Zealand white rabbits *Apolipoprotein E deficient mice

^ These near-infrared fluorescent magnetonanoparticles were used for detection of macrophage activity in atherosclerosis with laser scanning fluorescence microscopy.

in the gradient echo MR images of atherosclerotic rabbits, while no effect on control animals was observed (Hyafil et al., 2006; Ruehm et al., 2001; Yancy et al., 2005). Histology showed that macrophages were the dominant cell type for USPIO uptake; however, some endothelial cells and smooth muscle cells were occasionally also noted to contain iron oxide particles (Pande et al., 2006; Schmitz et al., 2000). Electron microscopy confirmed the uptake of USPIO in cytoplasm of macrophages (Ruehm et al., 2001).

Hyafil et al. (2006) showed that an indirect in vivo measure for USPIO-induced susceptibility artifacts in MR images of balloon-injured hypercholesterolemic rabbit aortas correlated well with RAM-11 positive areas (macrophages; r = 0.82, P < 0.05) and Perl's-positive areas (USPIOs; r = 0.78, P < 0.05), demonstrating that MRI can quantify uptake of USPIOs in macrophages. The indirect measure that was used in this study was the percentage of reduction in luminal areas, which is attributed to the extension of susceptibility artifacts outside anatomic borders of the vessel wall. Magnetic susceptibility artifacts could not be quantified directly in this study, because dark semicircular artifacts were observed at the water-fat interface in the frequency-encoding direction, also in the pre-contrast images. A high ex vivo correlation between signal intensity in MR images and iron content as quantified with inductively coupled plasma-mass spectrometry (ICP-MS) (r = -0.92, P < 0.001) was found by Yancy et al. (2005). In vivo quantification in the latter study was hampered by heterogeneity of plaque deposition, thickness of the MR image slice, and spatial in-plane resolution (Yancy et al., 2005). Schmitz et al. (2002) determined the accuracy for ex vivo MRI detection of USPIO accumulation in rabbit aorta vessel wall, using histology as standard of reference. They found a high accuracy, expressed as area under the receiver-operator characteristic (ROC) curve, of 0.85 and 0.88 for reader 1 and 2, respectively, with an interobserver agreement of 0.67. A drawback was the rather large number of false positive findings in 19% of the segments, partly caused by calcifications and mural thrombi. False-positive findings might be lower in the in vivo situation, when only those areas of new focal signal loss in post-contrast images are attributed to USPIO accumulation. Shrinkage of the histological specimens and subsequent matching problems of in vivo MR images with histology made in vivo analysis of accuracy impossible.

Different types and doses of USPIOs were used by the various groups (Table 5.2). The largest post-contrast signal decrease seemed to be observed by Ruehm et al. (2001) (Figure 5.3) and Hyafil et al. (2006) both using a very high dose of 1 mmol Fe/kg (« 20 times the normal clinical dose) and an USPIO agent with a high plasma half-life time (ferumoxtran-10, Sinerem®, Guerbet), allowing a larger time-window for the uptake of iron oxide particles by macrophages of plaque. Interestingly, in the first study conventional extracellular Gd-DOTA (Dotarem®) failed to reveal any abnormality in Watanabe hereditable hyperlipidemic (WHHL) rabbits, while abnormalities were clearly visible in the post-USPIO images (Ruehm et al., 2001).

The effect of USPIO dose was demonstrated by Schmitz et al. (2000), who observed a significant larger amount of areas of focal signal loss clearly confined

Figure 5.3. A Coronal MIP. B Sagittal oblique. C Coronal oblique reformatted images of contrast-enhanced 3D MRA data sets of a Watanabe hereditable hyperlipidemic (WHHL) rabbit obtained 5 days after intravenous injection of ferumoxtran-10. Note susceptibility effects originating within vessel wall and representing Fe uptake in macrophages embedded in plaque. Reprinted from Ruehm et al.(2001), with permission of Lippincott Williams & Wilkins.

Figure 5.3. A Coronal MIP. B Sagittal oblique. C Coronal oblique reformatted images of contrast-enhanced 3D MRA data sets of a Watanabe hereditable hyperlipidemic (WHHL) rabbit obtained 5 days after intravenous injection of ferumoxtran-10. Note susceptibility effects originating within vessel wall and representing Fe uptake in macrophages embedded in plaque. Reprinted from Ruehm et al.(2001), with permission of Lippincott Williams & Wilkins.

to the aortic wall on 31 ± 11 % of the MR images in the 200 ^mol Fe/kg group, while the number of images with this finding in the 50 ^mol Fe/kg groups were not significantly different from the control group without USPIOs. Histology confirmed this observation since a significant larger number of aortic segments of the high dose group were iron positive on a Prussian blue staining, while this was not the case for the low-dose groups. These findings may also partly be explained by a difference in post-contrast delay, which were 48 hours for the high-dose and 24 and 8 hours for the low-dose groups. A slight difference in histology findings was indeed observed between the groups with 24 and 8 hours delay, since in 21 ± 24% of the aortic sections of the first group a thin subendothelial rim of iron-positive cells was observed, while this was never observed in the latter group.

The distribution of USPIO in atherosclerotic plaque depends on the animal model. Yancy et al. (2005) demonstrated large differences in USPIO distribution in the aorta vessel wall of balloon-injured hypercholesterolemic New Zealand white rabbits, 2, 4, and 8 weeks post-injury. Ex vivo MRI (Figure 5.4) and histology (Figure 5.5) showed USPIO uptake in a rim in the intima along the internal elastic lamina 2 weeks post-injury, while a focal deposition of USPIO was observed in distinct areas throughout the intima and occasionally in media 4 and 8 weeks post-injury. Although the iron distribution was very different in these groups, ex vivo MRI signal decreases (^ 41 %-45%) were comparable.

Figure 5.4. Ex vivo MRI of the abdominal aorta. Representative single slice transverse gradient echo (a), proton density (b), T1 (c), and T2 (d) weighted spin echo images of the abdominal aorta from a 4 week post-injury control (-contrast) and 2, 4, and 8 week post-injury ferumoxtran-10 treated (+contrast) animals are shown. Images provided by courtesy of Jucker. Reprinted from Yancy et al. (2005), with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.

Figure 5.4. Ex vivo MRI of the abdominal aorta. Representative single slice transverse gradient echo (a), proton density (b), T1 (c), and T2 (d) weighted spin echo images of the abdominal aorta from a 4 week post-injury control (-contrast) and 2, 4, and 8 week post-injury ferumoxtran-10 treated (+contrast) animals are shown. Images provided by courtesy of Jucker. Reprinted from Yancy et al. (2005), with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.

In another animal model of 12-28-month-old WHHL rabbits, iron accumulation was predominantly observed in a rim on the luminal side of the aortic wall (Schmitz et al., 2000).

Based on cell morphology and RAM-11 staining, it appeared that USPIO accumulation was associated with smaller macrophages rather than larger foam cells (Yancy et al., 2005). Electron microscopy supported this observation, since macrophages containing cytoplasmic Fe particles were surrounded by foam cells without cytoplasmic iron (Ruehm et al., 2001). A disadvantage of electron microscopy is that only a very small sample is investigated which might not be representative.

A long plasma half-life time, allowing prolonged exposure of the vessel wall to USPIOs, is crucial for the capability of a certain type of USPIO to label atherosclerotic plaque. Yancy et al. (2005) compared the uptake of two different types of USPIOs with a large difference in plasma half-life time, ferumoxtran-10 and ferumoxytol, in rabbit atherosclerotic plaque. Rabbit plasma half-life as estimated from daily MR angiograms was approximately 8-fold higher for ferumoxtran-10. Although in vitro macrophage phagocytosis of ferumoxytol

Ferumoxytol

Figure 5.5. Histopathology of abdominal aorta from ferumoxtran-10 treated animals. Perl's iron (a-c) and RAM-11 immunohistochemistry (d-f) staining of abdominal aorta sections from 2, 4, and 8 week post-balloon injured rabbits, respectively. Iron accumulation was associated with macrophage in all sections (see arrows in each slide). Macrophage and iron were located primarily in the intimal (I) layer and only at the 8 week time point were macrophage detected in media (M) as seen in panel f. x20 magnification. L = lumen. Images provided by courtesy of Jucker. Reprinted from Yancy et al.(2005), with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.

Figure 5.5. Histopathology of abdominal aorta from ferumoxtran-10 treated animals. Perl's iron (a-c) and RAM-11 immunohistochemistry (d-f) staining of abdominal aorta sections from 2, 4, and 8 week post-balloon injured rabbits, respectively. Iron accumulation was associated with macrophage in all sections (see arrows in each slide). Macrophage and iron were located primarily in the intimal (I) layer and only at the 8 week time point were macrophage detected in media (M) as seen in panel f. x20 magnification. L = lumen. Images provided by courtesy of Jucker. Reprinted from Yancy et al.(2005), with permission of Wiley-Liss, Inc., a subsidiary of John Wiley & Sons, Inc.

was 4-6 fold greater than with ferumoxtran-10, ferumoxtran-10 was detected in vivo with MRI in all vessels while in vivo detection of ferumoxytol was negligible. With ex vivo MRI the uptake of ferumoxytol could also be visualized although the signal decrease was less than that in the ferumoxtran-10 groups. Iron quantification of in vitro plaque specimens with ICP-MS and histology (Perl's stain) confirmed the lower iron uptake in the ferumoxytol groups.

MRI reading is generally hampered by high USPIO concentrations in the lumen at low post-contrast imaging times. Schmitz et al. (2000) observed a circumferential area of low signal intensity which could not be attributed precisely to either the vessel wall or lumen at a post-contrast imaging time of 8 hours. This phenomenon is probably a susceptibility artifact due to the high USPIO concentration in the lumen. Indeed, no iron uptake was observed in corresponding histological sections (Schmitz et al., 2000). Similarly, Ruehm et al.. (2001) were unable to delineate the aortic wall on the first 2 days after contrast injection due to extensive susceptibility artifacts, and only after 4 days the wall could be delineated in all animals. Therefore, a sufficiently large post-contrast imaging time needs to be chosen.

In histology, the uptake of iron oxide nanoparticles is usually visualized indirectly, using a Perl's or Prussian blue staining. An interesting development is the use of near-infrared fluorescent iron oxide particles, which enables direct visualization of these nanoparticles in plaque with fluorescence miscroscopy (Pande et al., 2006).

Uptake of USPIOs causes a negative contrast (signal loss) in gradient echo images, which complicates image interpretation, since differentiation between USPIO uptake and dark-appearing areas like calcified tissue becomes laborious. Briley-Saebo et al. (2006) have shown in an initial study that a Gradient echo Acquisition for Superparamagnetic Particles (GRASP) sequence can generate positive USPIO-contrast in rabbits with balloon-injured hypercholesterolemic aortic wall. Signal enhancement observed using GRASP corresponded well with signal loss in the T2* weighted images. The value of this promising new pulse sequence for atherosclerotic plaque imaging remains to be validated in larger studies.

Was this article helpful?

0 0

Post a comment