of 37

In vivo Detection of Vulnerable Atherosclerotic Plaque by Magnetic. Resonance Imaging in a Rabbit Model

In vivo Detection of Vulnerable Atherosclerotic Plaque by Magnetic Resonance Imaging in a Rabbit Model Short title: Phinikaridou: In vivo detection of vulnerable plaque by MRI Alkystis Phinikaridou, PhD
0 views37 pages
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
In vivo Detection of Vulnerable Atherosclerotic Plaque by Magnetic Resonance Imaging in a Rabbit Model Short title: Phinikaridou: In vivo detection of vulnerable plaque by MRI Alkystis Phinikaridou, PhD * 1 ; Frederick L. Ruberg, MD 2,3 ; Kevin J. Hallock, PhD 4 ; Ye Qiao, PhD 1 ; Ning Hua; BSc 1 ; Jason Viereck, MD 5 ; and James A. Hamilton, PhD *1,6 1 Department of Physiology and Biophysics, Boston University School ol of Medicine, Boston, MA, USA 2 Section of Cardiology, ogy, Department of Medicine, Boston University it School of Medicine, Boston, MA 3 Department of Radiology, Boston University School of Medicine, Boston, MA 4 Department of Anatomy and Neurobiology, Boston University School of Medicine, Boston, MA, USA 5 Department of Neurology, Boston University School of Medicine, Boston, MA 6 Department of Biomedical Engineering, g Boston University Boston, MA * To whom all correspondence should be addressed 700 Albany Street, 3 rd Floor, W-302, Department of Physiology and Biophysics, Boston University School of Medicine, Boston, MA, Phone: Fax: Journal Subject Codes: [130] Animal models of human disease; [144] Other arteriosclerosis; [115] Remodeling; [172] Arterial thrombosis 1 Abstract: Background - The ability to identify atherosclerotic plaques with a high risk for sudden disruption prior to stroke or myocardial infarction would be of great utility. We used a rabbit model of controlled atherothrombosis to test whether in vivo magnetic resonance imaging (MRI) can noninvasively distinguish between plaques that disrupt after pharmacological triggering (vulnerable) and those that do not (stable). Methods and Results - Atherosclerosis was induced in male New Zealand White (n=17) rabbits by cholesterol diet and endothelial denudation of the abdominal aorta. After baseline (pretrigger) MRI with and without gadolinium contrast, the rabbits underwent two pharmacological triggerings to induce atherothrombosis, followed by another MRI 48 hours later (posttriggering). Atherosclerosis was identified by the pre-triggered images in all rabbits, and thrombosis was identified in 9/17 animals (53%) by post-trigger MRI. After sacrifice, 95 plaques were analyzed; 28 (29.5%) had thrombi (vulnerable) and 67 did not (stable) (70.5%). Pretriggered MRI revealed comparable stenosis in stable and vulnerable plaques, but vulnerable plaques had a larger plaque area (4.8±1.6 versus 3.0±1.0 mm 2 ; P=0.01), vessel area (9.2±3.0 versus. 15.8±4.9 mm m 2 ; P=0.01), and higher remodeling ratio (1.16±0.2 versus 0.93±0.2; P=0.01) compared to stable plaques. Furthermore, vulnerable plaques more frequently exhibited: (1) positive remodeling (67.8% versus 22.3%; P=0.01), in which the plaque is hidden within the vessel wall instead of occluding the lumen; and (2) enhanced gadolinium uptake (78.6% versus 20.9%; P=0.01) associated with histological findings of neovascularization, inflammation, and tissue necrosis. Conclusions - We demonstrate that in vivo MRI at 3.0 Tesla detects features of vulnerable plaques in an animal model of controlled atherothrombosis. These findings suggest that MRI may be used as a non-invasive modality for localization of plaques that are prone to disruption. Key words - magnetic resonance imaging, atherosclerosis, thrombosis, gadolinium, remodeling 2 Introduction Acute coronary syndromes (ACS) such as unstable angina pectoris and myocardial infarction are the leading causes of death in the United States 1. Histological studies demonstrate that ACS are usually triggered by rupture/erosion of vulnerable atherosclerotic plaques, which results in luminal thrombosis 2, 3. X-ray angiographic studies, which do not provide information about plaque composition, suggest that the majority of high risk plaques cause less than 50% luminal narrowing 4, 5. In vivo MRI can estimate the degree of luminal narrowing and identify plaque components 6-9. Contrast enhanced MRI (CE-MRI) using gadolinium-diethylenetriamine penta-acetic acid (Gd-DTPA) has improved the discrimination between the fibrous cap and the lipid core 10, 11 and necrotic core 12 and the visualization of coronary atherosclerosis e o 13, 14. Furthermore, dynamic CE-MRI (DCE) has shown that uptake of Gd-DTPA is correlated with neovascularization 15, 16 and inflammation 17, both of which are increased in vulnerable plaques. The inability to study thrombotic events and plaque vulnerability prior to atherothrombosis in humans necessitates the use of animal models. The rabbit model of controlled atherothrombosis using an intermittent cholesterol diet followed by pharmacological triggering was introduced by Contantinides et al. 18 and later modified by Abela et al. 19. Recently, we further modified the preparation/dietary protocol and demonstrated that such rabbits develop six out of eight types of plaques 20 classified by the American heart association (AHA) criteria. Importantly, we 20 and others 19, 21 have shown that pharmacologically induced thrombosis occurs in plaques with histological features of vulnerability. In an alternative rabbit model of atherothrombosis, plaques were ruptured after an inflatable balloon was embedded into the plaque 22. Although these rabbit plaques were histologically similar to human plaques, this model has not yet been used for MRI studies. In contrast, in vivo MR images of thrombosis 3 associated with plaque disruption in the rabbit model of pharmacologically induced thrombosis have been reported 23, 24. In this study, we used the rabbit model of experimentally induced atherothrombosis 20 to explore whether MR images obtained in vivo at 3.0 Tesla could identify plaques prone to disruption. Methods Animal model Atherosclerosis was induced in adult male New Zealand white rabbits (n=24, ~2.8 kg, Charles River Laboratories, MA) as previously described 20. Briefly, rabbits bits were fed a 1% cholesterol diet (PharmaServe, MA) for 2 weeks prior to and 6 weeks after balloon injury of the abdominal aorta, followed lowed by 4 weeks of normal chow diet (Figure 1). Balloon injury of the aortic wall was performed under general anesthesia [acepromazine (0.75 mg/kg IM), ketamine (35 mg/kg, IM) and xylazine (2.5 mg/kg, IM)]. Pharmacological triggering of thrombosis was induced with Russell s viper venom (0.15 mg/kg IP; Enzyme Research, IN); an activator of Factor X of the coagulation cascade followed 30 min later by histamine; a vasoconstrictor in rabbits (0.02 mg/kg IV; Sigma-Aldrich, MO). This procedure was performed twice, within 48 h, in each animal, as previously described 18-21, 23, 24. Within 24 h after the post-trigger MRI, the rabbits received heparin (1000 USP units IV, Sigma-Aldrich) to prevent post-mortem blood clotting, and were sacrificed with a bolus injection of sodium pentobarbital (100 mg/kg IV). Subsequently, the aortas were excised and fixed in 10% formalin for histological analysis. Three age- and gender-matched, uninjured rabbits were fed only normal chow diet and used as controls. 4 Of the 7 rabbits that did not complete the study, 3 died prematurely from respiratory distress, ischemic heart disease, and/or liver failure (data not shown), 2 became anorexic early in the experimental protocol and were returned to normal diet, and 2 rabbits became paralyzed from the waist down after the first pharmacological triggering and were euthanized before the end of the protocol. Histological analysis of these 2 rabbits revealed occlusive thrombosis in the distal aorta (data not shown). Animal studies were performed in accordance with guidelines approved by the Institutional Animal Care and Use Committee of Boston University. MRI experiments In vivo MRI experiments were performed on supine rabbits under deep sedation using a 3.0 Tesla Philips Intera Scanner (Philips Medical Systems, OH) and a synergy knee coil with six elements. A pulse oximeter was placed on the ear for cardiac gating. The aorta of atherosclerotic rabbits was imaged before (pre) and 48 h after (post) the first pharmacological triggering (Figure 1). Control rabbits were imaged once. MRI acquisition parameters are listed in Table 1. Un-gated coronal 3D phase contrast MR angiograms (PC-MRA) acquired with a T1-weighted, fast-filed echo sequence were used as scout images. Then 2D T1-weighted black-blood (T1BB) axial images (4 mm) were acquired with a double inversion recovery turbo spin echo sequence and cardiac gating (at every other systolic phase). Subsequently, un-gated axial 3D PC-MRA images were acquired immediately after a bolus injection of Gd-DTPA (0.1 mmol/kg, IV) (Magnevist, Germany). For every axial T1BB slice (4 mm), eight 0.5mm PC-MRA slices were acquired. Finally, post-contrast enhanced (post-ce) T1BB images were acquired min after Gd- DTPA injection with parameters identical to those used for the non-contrast enhanced T1BB images. 5 Plasma lipid and inflammatory marker analysis Blood samples were collected after overnight fasting from the ear artery at baseline, at the end of the 8-week cholesterol diet, and before triggering. Plasma total cholesterol (TC) and HDL-cholesterol (HDL-C) were measured with enzymatic reaction kits from BioVision (Mountain View, CA) and Wako Chemicals Co. (Richmond, VA), respectively. C-reactive protein (CRP) and plasminogen activator inhibitor-1 (PAI-1) were measured with ELISA kits from Immunology Consultant Laboratory (Newberg, OR) and Molecular Innovations (Southfield, MI), respectively. Matching of MR images and histological sections The distances s from the aortic renal branches and the iliac bifurcation were used as internal anatomical markers to match the MR images and histological sections 20. Histology and identification of vulnerable plaques Transverse cryo-sections (10 μm) were collected throughout the length of the each segment and stained with Masson s trichrome (Sigma Aldrich) to identify cellular components and thrombi. Disrupted (vulnerable) plaques were defined as those with attached platelet and fibrin-rich thrombi. We have previously demonstrated that thrombosis in this model originates both from rupture of thin cap atheromas (60%) and superficial plaque erosion (40%), which frequently occurred over plaques classified by the AHA as atheromas and fibroatheromas, and rarely over fibrotic plaques 20. Plaques that had no overlying thrombus were defined as nondisrupted (stable). 6 Analysis of MR images Because the rabbits developed plaques throughout the region of the aorta that was balloon injured, we performed a slice-by-slice analysis of axial wall images containing stable and vulnerable atherosclerotic plaques. Of the 204 pre-triggered T1BB slices acquired from 17 rabbits, a total of 190 slices, (95 before and 95 after gadolinium administration), and 1520 PC- MRA slices (760 anatomical and 760 flow-encoded) slices were evaluated. 109 T1BB images were excluded from the analysis: 58 because either the T1BB or the PC-MRA slices were of poor image quality (insufficient blood suppression, bad signal-to- noise, motion artifacts) and 51 because the corresponding PC-MRA slices contained side branches that could impair the assessment of the remodeling ratio. Aortic regions containing plaques detected ecte ted in the pretriggered MR images were classified into stable and vulnerable, based on the presence of luminal thrombosis seen on the post-triggered T1BB images and the corresponding histopathology. Subsequently, the only the pre-triggered MR images were analyzed using ImageJ (NIH). Pre-CE T1BB images were used to calculate the plaque area (PA) and the % crosssectional narrowing (CSN) by manually segmenting the adventitial and luminal contours of the vessel wall. Plaque area was calculated as: PA = adventitial area - lumen area and the CSN as % CSN = (plaque area/vessel area)*100. Un-gated 3D PC-MRA images acquired immediately after injection of Gd-DTPA were used to calculate the remodeling ratio (RR) and the % stenosis from flowcompensated/anatomical and flow-encoded images, respectively. In the anatomical images (T1- weighted spoiled-gradient echo) flowing blood appears bright whereas the contrast of stationary tissues depends on the T1 relation times. In flow-encoded images, only flowing spins elicit 7 signal, and the intensity is proportional to the velocity of flow, whereas stationary tissues are suppressed. It has been shown that spoiled-gradient echo images detect the adventitia/outer region of the vessel wall and that the delineation of this contour becomes improved in contrastenhanced images A comparison of the vessel area measured on different MRI images is shown in Table 1 (supplemental data). Thus, at each lesion site, the anatomical images were used to measure the vessel area (VA) for the calculation of the RR, and the corresponding flowencoded images were used to calculate the unobstructed lumen area (LA) and the % stenosis. The RR and the % stenosis were calculated after correcting for arterial tapering 28 and interindividual variability of arterial size 29. The RR was calculated as RR = vessel area lesion/ vessel area reference (Figure 3A) and the three remodeling categories were defined ed as previously described 30 : positive if RR 1.05, intermediate if 0.95 RR 1.05 and negative if RR 0.95. The % stenosis was calculated as: % stenosis=1-[lumen area a lesion/lumen area reference r ]*100. Because of diffuse vessel wall thickening, the slice with the least amount of plaque was used as a reference site, assuming that it was least affected by the disease (mean values of references: PA = 2.0±0.56 mm 2, VA = 11.0±3.5 mm 2, LA = 7.2±1.5 mm 2 and % CSN = 21.4±6.3). Post-CE T1BB images were visually compared to the pre-ce T1BB images to evaluate the presence or absence of a circumferential (full ring) or crescent-shape enhancement pattern of the vessel wall. Bright signal from perivascular lymphatics and/or adipose tissue was sometimes visible in the pre-ce T1BB images. To eliminate ambiguities in the evaluation of the enhancement pattern of gadolinium-enhanced images, these regions were outlined on the pre-ce T1BB images and subsequently masked onto the gadolinium-enhanced images. Statistical Analysis 8 Analyses were performed using SPSS 11.0 (SPSS Inc). For 2-group comparisons, continuous variables were compared using either a two-sample t-test or a Mann-Whitney nonparametric test after the variables were ranked. Categorical variables were compared using the 2 test. Qualitative data are presented as frequencies. Two independent observers (A.P and J.V.) analyzed the pre-ce T1BB images to calculate the plaque area and evaluated the enhancement pattern on the post-ce T1BB images. In addition, two independent observers (A.P. and N.H) analyzed the PC-MRA images to calculate the vessel and lumen areas. Observers (J.V and N.H) were blinded to the MRI and histological findings. The inter-observer variability was assessed by using the inter-class correlation coefficient (ICC) for continuous variables and Cohen s kappa for categorical variables. Independent predictors of plaque vulnerability were identified by multi-logistic regression analysis after the plaques were categorized as vulnerable and stable. Variables exhibiting statistical significance in the univariate regression (i.e., plaque area, vessel area, remodeling index, presence of gadolinium hyper-enhancement, presence of positive and negative remodeling) were then used in the multi-logistic regression model. Multiple linear regression analysis was used to evaluate the relationship between plasma biomarkers and plaque vulnerability. The sensitivity, specificity, positive and negative predictive values (PPV and NPV), and diagnostic accuracy of the MRI features alone or in combination were calculated. Data are presented as mean ± SD. Probability values of P 0.05 were considered significant. Results Atherosclerosis and thrombosis can be imaged by MRI 9 Aortic plaques were located in vivo using the pre-triggered MR images, and the sites of luminal thrombosis were visualized on MR images acquired 48 h after pharmacological triggering. The sites of plaques, plaque disruptions, and thrombosis were validated by the corresponding histological sections. Atherosclerosis was observed in all rabbits and thrombosis occurred in 9/17 (53%) of them. No atherosclerosis was observed in control rabbits. A total of 95 wall segments containing plaques were included in this study, of which 28 (29.5%) showed luminal thrombi and 67 did not (70.5%). Figure 2 shows representative MR images and histopathology of a plaque that did not disrupt (Figure 2A-C) and a plaque that disrupted after triggering g (Figure 2D-F). The pretriggered image of the stable plaque (Figure 2A) demonstrates an eccentric entr plaque that did not change in appearance e after pharmacological triggering (Figure 2B). The corresponding o histological section confirmed the presence of an intact fibrous cap overlaying a lipid-core (Figure 2C). The pre-triggered image of the vulnerable plaque (Figure 2D) shows the plaque, whereas the post-triggered image shows a new mass protruding into the lumen (Figure 2E). The corresponding histology (Figure 2F) revealed the site of plaque rupture and confirmed the presence of an overlying platelet- and fibrin-rich thrombus. Quantitative MRI and MRA measurements of stable and vulnerable plaques Plaque area, vessel area and remodeling ratio were significantly larger in vulnerable plaques; however, luminal area, % stenosis and % CSN were similar between the two groups (Table 2). ICC revealed a high inter-observer agreement for the measurements of vessel area (ICC = 0.92, 95% CI = ) and luminal area (ICC = 0.9, 95% CI = ), and 10 moderate inter-observer agreement for the measurement of plaque area (ICC = 0.64, 95% CI = ). Vulnerable plaques are associated with positive remodeling A key finding from the analysis of the pre-triggered images is that the plaques that disrupted after pharmacological triggering frequently exhibited positive remodeling (Figure 3). Pre-triggered PC-MRA images acquired from the same rabbit demonstrate examples of negative and positive remodeling compared to a reference site. The vessel area measured at the site of the stable plaque (Figure 3B) was smaller than that of the reference site (Figure 3D), which is indicative of negative remodeling. In contrast, the vessel area of the vulnerable plaque (Figure 3F) was markedly larger than the reference site, suggestive of positive remodeling. Images of the lumen (Figure 3C, E, G) demonstrate that this example of a stable plaque exhibited a greater extent of stenosis compared to that calculated for the vulnerable plaque. Overall, stable plaques frequently exhibited negative remodeling whereas vulnerable plaques frequently exhibited positive remodeling (Figure 3H). Similar findings were obtained when the frequency of the remodeling types in stable and vulnerable plaques was calculated using T1BB and post-ce T1BB images (Table 2; supplemental data). Vulnerable plaques show hyperintense enhancement after administration of Gd-DTPA Another key finding that emerged from the analysis of the pre-triggered MR images is the hyperintense signal associated with vulnerable plaques after administration of Gd-DTPA. A stable plaque that showed mild uptake of Gd-DTPA (Figures 4A, B) had a thick fibrous cap 11 overlaying a lipid-core (Figure 4C). In contrast, vulnerable plaques demonstrated hyperintense circumferential (Figure 4E) or crescent-shaped enhancement (Figure 4H) that extended beyond the plaque. The corresponding histology (Figure 4F, I) confirmed that a thrombus formed after pharmacological triggering and revealed extensive neovessels, in the fibrous cap (Figure 4F; arrow), the intima and the adventitia (Figure 4I; circles). Furthermore, histology revealed degradation of the extracellular matrix and tissue necrosis, two additional contributors of increased gadolinium uptake. Contrast enhancement of the vessel wall was not observed in
View more
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!