Background:

It is now widely questioned that atherosclerotic plaques with hemodynamically significant stenosis cause the majority of acute myocardial infarction and stroke. Rather, they are caused by the so-called "vulnerable plaques". These plaques, retrospectively characterized by large lipid cores, thin fibrous caps, dense superficial macrophage infiltration, or endothelial denudation with thick cap, are prone to induce thrombosis and sudden luminal occlusion. It is still impossible for existing diagnostic techniques to accurately predict which plaque is going to cause luminal thrombosis.

We believe the answer may not be found solely in the structural evaluation of the plaques. In other words, atherosclerotic lesions with similar structural features may not behave similarly. The need for assessment of functional properties or activity of plaques, in particular their monocyte recruitment rate and macrophage activity, new angiogenesis, matrix proteolysis has led us to pursue “functional assessment” of plaques.

We believe Magnetic Resonance Imaging (MRI) has the potential to fulfill all the above in addition to illustrating the anatomy of plaque and lumen.

We have sought a method to detect inflammation (macrophage density), leaking angiogenesis and fissured/permeable cap of atherosclerotic plaques based on their uptake of nano-particles of Super Paramagnetic Iron Oxide (SPIO).

SPIO:SPIO and ultra-SPIO (USPIO) have a central core of iron oxide generally coated by a polysaccharide layer. These nano-particles are taken up 10-100 times more by macrophages than other cells, and also leak out through loose endothelial junctions of new vessels. On MRI, they shorten relaxation time by 10 folds or more and produce a sharp dark contrast by virtue of signal reduction.

Here, we present our preliminary findings on detection of atherosclerotic plaques in atherosclerotic rabbits using SPIO nano-particles.

Hypothesis:

We hypothesized that certain features associated with plaques vulnerability (i.e. Inflammation, angiogenesis, intra-plaque hemorrhage, and fissured/permeable cap) may cause higher uptake of SPIO by atherosclerotic plaques compared with normal arterial wall.

Fig 1. Introducing a novel method for MR imaging of atherosclerotic plaque to identify plaque inflammation, angiogenesis, vasa vasorum, fissured and permeable cap.

“SPIO Effect” SPIO-induced decreased signal intensity is not proportional to the size of SPIO. In other words, SPIO particles produce a big dark halo around them, much larger than their actual size (over-magnification), specially in T2 images. The above schematic figures (Fig.1) represent a vulnerable plaque taking up more SPIO compared to a stable plaque. Non-stenotic, yet vulnerable plaques do not show luminal narrowing in ordinary MRI or MRA. However, after injecting SPIO, a plaque loaded with SPIO can be detected as a big dark spot along the arterial line, as if there is a stenotic plaque obstructing blood flow. This phenomenon may be called “SPIO Effect”.

METHODS:

In order to study the distribution of iron in different tissues, 3 WHHL rabbits and 2 NZW rabbits (controls) were injected with SPIO (2 mmol Fe/kg) IV through an ear vein. One WHHL and one NZW rabbit served as untreated controls (i.e., they received no SPIO). Animals were sacrificed on postinjection day 5 and 10. Tissues from the aorta as well as liver, spleen, kidneys, and heart were fixed and stained for H&E, iron, and RAM 11 (rabbit anti-macrophage antibody).

In Vivo:

Using a 1.5T MRI system (Signa, General Electric) equipped with a conventional extremity coil, baseline MRI of the aorta was done in 4 WHHL and 2 NZW rabbits (T2 gradient echo: TR = 1200 msec, TE = 6 msec, FOV = 16 x 12 cm, matrix size = 256 x 192 pixels; 3-dimensional magnetic resonance [3D MR] angiography with gadolinium-DTPA: TE = 1.3 msec, TR = 5.6 msec). The rabbits were injected with SPIO (2 mmol Fe/kg) IV via an ear vein. Post-contrast MRI was performed on day 5 using the same MRI sequences. MRI was done with respiratory and cardiac gating. The rabbits were anesthetized with isoflurane for the duration of their studies.

Ex Vivo:

All rabbits that underwent in vivo MRI were sacrificed. The aortas were excised, isolated, and placed in a gel medium. Both ends of the aorta were clamped and all side branches were occluded. Gadolinium-DTPA was injected inside the lumen. Then, MRI was performed, using the 1.5T scanner used in the in vivo experiments (Signa, General Electric). Data on T2 gradient echo and 3D MR angiography sequences were recorded for each specimen.

RESULTS:Histopathologic studies in rabbits revealed accumulation of iron in the atherosclerotic arterial wall (Fig. 2). The correlation between iron accumulation and macrophage accumulation in the aortic wall was significant (r = 0.95). Actively inflamed atherosclerotic areas of the aortic wall showed higher uptake of SPIO than did the normal aortic wall (RAM-11 positive) and noninflamed atherosclerotic areas. SPIO particles were not evenly distributed in all plaques. Areas with thick fibrous caps accumulate less SPIO while areas with minimal fibrosis and an abundance of subendothelial foamy cells accumulate more. Electron microscopy studies showed that almost all SPIO particles were intracellular (Fig. 3). They also revealed sporadic localization in endothelial cells, though this may simply indicate diffusion into permeable areas of the endothelium.

Fig. 2. Histopathology of the aortic wall in hypercholesterolemic (WHHL) and normal (NZW) rabbits. Shown are the aortic wall in a WHHL rabbit (A-C) and a NZW rabbit (J-L) 5 days after SPIO injection. Also shown are the aortic wall in a WHHL rabbit (D-F) and a NZW rabbit (G-I) serving as untreated controls. Staining was done for hematoxylin and eosin (panels A, D, G, and J), iron (panels B, E, H, and K), and rabbit anti-macrophage antibody (RAM11) (panels C, F, L, and I). Magnification 10x for panels A, D, G, and J; magnification 40x for all other panels.

Fig. 3. Electron microscopy of the aortic intima in a WHHL rabbit, revealing iron particles inside foamy macrophage cells (left) and inside an endothelial cell (right). Magnification 8000x for left panel; magnification 2400x for right panel.

WHHL vs. NZW Rabbits

In vivo MRI studies revealed decreased signal intensity on 3D MR angiography in the aortic wall (Fig. 4).

Fig. 4. In vivo images of the aorta in atherosclerotic (WHHL) and normal (NZW) rabbits, obtained by 3-dimensional (3D) TOF magnetic resonance angiography with gadolinium-DTPA before and 5 days after SPIO injection. (A) WHHL rabbit before injection; (B) WHHL rabbit after injection; (C) NZW rabbit before injection; (D) NZW rabbit after injection.

Because SPIO uptake in the liver, spleen, bone marrow, and other tissues imposed a tissue artifact, changes in T1- and T2-weighted images of the aorta or other arteries could not be appreciated.

Ex vivo MRI studies were done to negate the effect of the tissue artifact mentioned above. As revealed by 3D MR angiography, there were significant luminal irregularities in the aortic walls of SPIO-injected WHHL rabbit. Also, as shown by T2*- weighted images of the SPIO-injected WHHL rabbit, SPIO had a negative enhancement effect in the atherosclerotic aortic wall (Fig. 5).

Fig. 5. (Top) Ex vivo images of the intraaortic lumen in atherosclerotic (WHHL) and normal (NZW) rabbits, obtained by 3-dimensional (3D) TOF magnetic resonance angiography with gadolinium-DTPA. (A) WHHL rabbit injected with SPIO; (B) WHHL rabbit not injected with SPIO; (C) NZW rabbit injected with SPIO; (D) NZW rabbit not injected with SPIO. (Bottom) T2 gradient echo magnetic resonance imaging sequences for the same subjects, in the axial view (E-H, respectively).

Conclusion:

Histologic examination of SPIO injected in WHHL rabbits showed a significantly higher uptake of SPIO particles by aortic atherosclerotic lesions than normal arterial wall (both within the same animals and also compared to NZW rabbits). These particles can be found in the plaque as early as 3 hours (data not shown), and as late as 10 days post injection.

MR imaging of the rabbit aorta both in-vivo and ex-vivo revealed the reduction in signal intensity in the aorta after injection of SPIO. This effect could be seen mainly in T2* and 3D angiogram sequences.

Implications :Our preliminary findings may have clinical application in detection of vulnerable plaques using MRI. The goal should be to achieve plaque-targeted SPIO (i.e. ox-LDL and ICAM-1 antibody-conjugated SPIO, under development in our laboratory). This goal holds promise for more precisely locating vulnerable plaques by MRI. Since monocyte/macrophage system is the major source for SPIO accumulation in the plaques, histologic studies of atherosclerotic plaques after the injection of SPIO could help us study on the dynamics of macrophage movement in and out of the plaque. Further information on inhibitors and stimulators of macrophage homing could be obtained using this novel macrophage tracer in the near future.