A Novel Nonobstructive Intravascular MRI Coil
In Vivo Imaging of Experimental Atherosclerosis
Objective— MRI is being used to characterize the composition of atherosclerotic plaques. However, the resolution achievable using surface radiofrequency coils is limited by the signal to noise ratio. We studied the utility of a new intravascular (IV) MRI probe for high-resolution in vivo imaging of atherosclerotic lesions.
Methods and Results— Balloon-injured Watanabe heritable hyperlipidemic rabbits served as the experimental model of atherosclerosis. The newly developed IV MRI probe is 1.3 mm in diameter and can be positioned over a guidewire. MRI was performed with both an external phased-array coil and the IV MR coil. MR observations were correlated with histopathology. After MRI, the animals were killed and analysis of agreement between MR and histopathology was performed. The IV MR coil allows aortic images to be obtained with 156×156 μm2 in-plane resolution versus 352×352 μm2 when used with the external phased-array coil. No significant motion artifacts were noted, despite the continuation of arterial blood flow during image acquisition around the IV probe. The different components of the atherosclerotic lesions (lipid core and fibrous cap) were easily identified. There was an excellent agreement between MRI with the IV coil and histopathology by simple linear regression for both the mean wall thickness (r=0.88, slope 0.82, P<0.0001) and vessel wall area (r=0.86, slope 1.08, P<0.0001).
Conclusions— The new nonobstructive design for this intravascular coil provides great promise for additional work in high-resolution MRI characterization of atherosclerotic plaques in vivo. The ability to position the probe with a guidewire allows its placement under fluoroscopic or MRI guidance, whereas its size is compatible with human coronary arteries.
The pathogenesis of the acute coronary syndromes is frequently related to atherosclerotic plaque disruption and subsequent thrombosis. Atherosclerotic plaque composition, rather than the degree of arterial stenosis, seems to be a critical determinant of both risk of rupture and subsequent thrombogenicity.1 In particular, a large lipid core and a thin fibrous cap render a lesion susceptible or vulnerable to disruption and thrombosis. Modification or stabilization of vulnerable plaques in the coronary arteries, by strengthening the fibrous cap and decreasing the lipid core, has been proposed as an important mechanism responsible for the observed beneficial clinical effect of this lipid lowering.2 Thus, an imaging modality able to define the composition of the atherosclerotic lesions could potentially not only allow the identification of these vulnerable atherosclerotic lesions but also monitor effects of therapeutic interventions for stabilization on the plaque composition.
MRI allows the detection and characterization of atherosclerotic lesions in humans and animal models of atherosclerosis.3 We have recently shown that in vivo noninvasive MRI can identify lesions in rabbit models of atherosclerosis.4,5⇓ However, the combination of cardiac and respiratory motion artifacts, nonlinear course, and relatively small size of the coronary arteries handicap the ability of MR to accurately identify and quantify atherosclerotic plaques in the human coronary arteries. New MRI sequences for the coronary artery wall in animal models6 and humans7,8⇓ have been reported, overcoming the issues of motion suppression. However, additional improvements in resolution are required before robust in vivo documentation of atherosclerotic plaque burden and composition can be performed in the human coronary arteries in vivo.
Intravascular (IV) MR coils are a potential method of enhancing the signal to noise ratio, thus improving the resolution for arterial wall imaging. To date, reports of atherosclerotic imaging with an IV MR coil have required the use of occlusive balloons to stabilize the coil9,10⇓ or been performed postmortem.11 We have designed a novel IV MR coil that affords a high signal to noise ratio at the level of the artery wall without occluding the artery lumen.
In this study, we report the utility of a novel IV MR coil that permits high signal to noise ratio for the in vivo imaging of atherosclerosis in the Watanabe heritable hyperlipidemic (WHHL) rabbit without occluding the artery lumen.
Animal Model of Experimental Atherosclerosis
In this study, WHHL rabbits (n=5, age 3 months, weight 2.5 to 3.0 kg) underwent denudation of the aorta from the renal arteries to the iliac bifurcation 1 week after arrival to the institution, as previously described.4,5⇓ The endothelial denudation combined with the high cholesterol levels typical of this animal model results in the induction of atherosclerotic-like lesions. All imaging and procedures were performed under general anesthesia by intramuscular injection of ketamine (Fort Dodge Animal Health) (20 mg/kg) and xylazine (Bayer Corporation) (10 mg/kg). The Mount Sinai School of Medicine animal management program, under accreditation from the American Association for the Accreditation of Laboratory Animal Care, approved all interventional procedures.
IV and Surface Coil MRI
All rabbits had MRI performed 6 months after aortic balloon denudation using a 1.5 Tesla MRI system (Signa CV/I GE Medical Systems) with enhanced gradients (40 mT/m) and slew rates (150 T/m/sec). MR images were obtained with both the IV MR coil and a conventional phased-array volume coil under general anesthesia, as described above. The IV MR coil is based on a single-loop receiver coil design and includes a lumen for a 0.014-inch diameter guiding wire; it was manufactured by Magna-Laboratory Inc and should be commercially available within 1 year. The IV MR coil is 1.3 mm in diameter and 57 mm long (4.3F) and can be positioned over a guidewire (Figure 1). The IV coil was initially tested in a phantom filled with 30 mmol/L nickel chloride solution. The phantom has a 2-mm tube in the center, into which the catheter was placed (Figure I, available online at http://atvb.ahajournals.org). A study of the radial signal to noise ratio characteristics was produced using standard NEMA procedures. The signal to noise ratio was substantially higher within a 5-mm radius of the coil. The coil was interfaced to the GE CV/I scanner by use of an inductive Pi impedance mating circuit. The Pi circuit is tuned for an output impedance of 50 ohms (Figure II, available online at http://atvb.ahajournals.org). This in turn is plugged into the GE surface coil input box on the MRI scanner.
For in vivo studies, the IV MR coil was inserted via the right femoral artery using a sterile technique. The IV MR coil imaging was performed first. Sequential axial images (3-mm thickness) of the aorta from the renal arteries to the iliac bifurcation were obtained using a fast spin-echo sequence using the following parameters: PDW:TR/TE, 2300/17 ms; T2w, TR/TE:2300/60 ms; field of view, 3×3 cm; matrix, 256×256; echo train length, 8, signal averages, 4. Thus, the in-plane resolution was 156×156 μm2. These parameters were preselected on the basis of preliminary informal ex vivo work, confirming the feasibility of these parameters. Furthermore, this resolution is comparable to that achieved by published studies by other groups.9–11⇓⇓
Then, without moving the rabbit in the magnet, the IV MR coil was removed and the exact same imaging sequences were performed using the phased-array coil, with the only difference being that the field of view was 9×9 cm, providing an in-plane resolution of 352×352 μm2. Therefore, the MRI time was the same for both protocols, ≈10 to 15 minutes, respectively. Flow saturation pulses were used as previously described.4,5,12,13⇓⇓⇓ Therefore, exact image matching could be performed and confirmed by using aortic branches such as the renal arteries and the iliac bifurcation to allow exact coregistration, as previously described.4,5⇓
The MR images were transferred to a Macintosh computer for additional analysis. The MR images from both the IV MR coil and phased-array coil were matched with each other and with the corresponding histopathological sections for the aortic specimens, as described above. Cross-sectional areas of the lumen and outer boundary of each aortic section were determined using manual tracing with Image Pro-Plus (Media Cybernetics) by an observer blinded to the results of the histopathology. The manual tracing of the MR images was performed by an independent observer, experienced in vessel wall tracing, blinded to the results of the histopathology (S.G.W.). From these measurements, lumen area, outer vessel area, and vessel wall area (outer vessel area minus lumen area) were calculated. The outer wall was defined as the vessel wall–epicardial fat interface. This technique has been extensively used by our group and well validated previously, with excellent intraobserver and interobserver variability reported.5,6⇓
Euthanasia was performed immediately on completion of MRI of the rabbits by intravenous injection of Sleepaway 5 mL IV (Fort Dodge Animal Health) after receiving heparin (100 U/kg) to prevent postmortem blood clotting. The aortas were immediately flushed with 250 mL of physiological buffer (0.1 mol/L PBS, pH 7.4) followed by perfusion fixation with 250 mL cold (4°C) 4% paraformaldehyde in 0.1% PBS. After perfusion fixation, all specimens were immersed in fresh fixative and stored at 4°C. Coregistration was performed carefully by using the position of the aortic arch, renal arteries, and iliac bifurcation. Aortic specimens were embedded in paraffin, and sections 5-μm thick were cut and stained with combined Masson’s trichrome elastin stain.
The histopathological sections were digitized to the same computer and the same parameters, analyzed with the method described above for MR image analysis. The manual tracing of the histopathology was performed by an independent observer experienced in vessel-wall tracing, blinded to the results of the MR images (G.H.). The outer wall of the vessel was defined as the dense adventitial-epicardial fat interface on histopathology. Investigators blinded to the results of MRI performed analysis of these segments.
Comparisons between IV MR coil images and histopathology assessment of vessel parameters in the validation subgroup were performed using simple linear regression analysis with 95% confidence intervals (Statview, SAS Institute Inc) and the Bland-Altman method.14 To account for repeated observations within the same animal, fixed-effects models were estimated for vessel wall area and mean wall thickness. In each model, the MR value of the measurement was the dependent variable and the histology measurement was the independent variable. The fixed-effects model permits estimation of a constant (across animals) slope relating the two measurements but allowing a different intercept for each animal. Because the fixed effects model found heterogeneity of intercepts among the 3 animals for both lesion attributes, we then estimated a model in which each animal’s data were permitted to have a separate slope as well as a separate intercept and then tested for heterogeneity of slopes. Calculations were performed using Stata version 7.0 (Stata Corp).
IV Versus Surface Coil MRI
The newly designed IV MR coil allowed high-resolution images with an in-plane resolution of 156×156 μm2 to be reproducibly obtained. Importantly, there were no significant motion artifacts detected despite the arterial flow of blood around the catheter. The presence of circulating blood is confirmed by the dark appearance of the aortic lumen on the MR images with the flow suppression MR techniques, and one can appreciate the minimal obstruction to the aortic lumen by the IV coil (Figure 2). Furthermore, we could clearly distinguish the lipidic from the fibrotic regions in the atherosclerotic plaques using T2W imaging, as previously validated.15,16⇓ However, using the phased-array MR coil, an in-plane resolution of 352×352 μm2 only could be obtained (Figure 2).
Correlation of IV MR Coil and Histopathology
We compared matched IV MR coil images with the corresponding histopathological sections of the abdominal aorta (n=30) (Figure 3). There was a good agreement between IV coil MRI and histopathology for the measurement of both vessel wall area (r=0.86, slope 1.08, P<0.001) and mean wall thickness (r=0.88, slope 0.82, P<0.001), as assessed by simple linear regression analysis (Figure III, available online at http://atvb.ahajournals.org). This was confirmed by review of the Bland-Altman analyses, with a mean difference (IV coil MRI minus histopathology±SD) for vessel wall area of 0.89±0.54 mm2 and for mean wall thickness of 0.00±0.05 mm (Figure IV, available online at http://atvb. ahajournals.org). Thus, there was a tendency for measurements by MRI to be slightly larger than by histopathology, a phenomenon previously described.4,10,15⇓⇓ In the fixed-effects regression models, the within-rabbit coefficient of determination (R2) for vessel wall area was 0.78, and for mean wall thickness, 0.67. The respective slopes were 0.96 and 0.71. For both vessel wall area and mean wall thickness regressions, the hypotheses that the intercepts are equal for all 3 rabbits were rejected (P<0.00005 and P=0.0049, respectively). In an expanded model in which, effectively, separate regression lines are fitted for each rabbit, we tested the hypothesis of equality of slopes across rabbits. These hypotheses were not rejected: P=0.94 for vessel wall area and P=0.52 for mean wall thickness. Accordingly, we accepted the fixed-effects regression (heterogeneous intercepts with common slope) as the final model for the relationship between MR and histology measurements of both vessel wall area and mean wall thickness.
The new nonobstructive in vivo IV MR coil has significantly improved the resolution obtained over conventional surface phased-array coil imaging. This improvement allows the accurate detection and characterization discrimination of aortic atherosclerotic lesions in this experimental model. Given the critical role of atherosclerotic plaque composition in lesion vulnerability and subsequent thrombogenicity, the availability of an imaging device able to clearly define plaque composition would have significant clinical implications. Catheter-based imaging techniques may be required to allow adequate resolution for the accurate quantification and characterization of human coronary atherosclerotic lesions. In this regard, the possibility of acquiring images without totally occluding the arterial lumen facilitates the use of this type of device for the imaging of human coronary arteries. We have shown that a nonobstructive IV MR receive coil is able to improve the resolution obtained with surface coil imaging for atherosclerotic imaging.
Several studies have now demonstrated in vivo MRI of noncoronary atherosclerotic lesions in humans.16–19⇓⇓⇓ However, it is acknowledged that substantial improvements in the signal to noise ratio and reductions in the significant motion artifacts would be required before coronary atherosclerosis imaging could be performed. Clearly, additional studies are required to define the ability of this IV MR coil to image atherosclerotic lesions in the coronary arteries.
The small but consistent overestimation of mean wall thickness and vessel wall area by MRI in comparison to histopathology may relate to partial volume effects by MRI as well as shrinkage of the histopathology specimens as a consequence of their preparation.4,10,15⇓⇓ Our regression models and graphs show that for any given animal, there is a strong linear correlation between the vessel wall area as measured by MR and the same lesion size as measured histologically. The same conclusion holds for mean wall thickness as well. The Bland-Altman analyses show that in general there is a good agreement between MR and histopathological measurements, although additional work is needed to improve the accuracy of MRI for the quantification of coronary lesions.
We selected the WHHL rabbit for this study because it has an its intrinsic deficiency in LDL receptors, thus more closely approximating humans with atherosclerosis than the cholesterol-fed rabbit models.20 Furthermore, we have reported the utility of external MRI in this model for detecting changes in vessel wall dimensions, allowing the determination of vessel wall remodeling.4 However, in that study, we were also unable to characterize the atherosclerotic lesions of the WHHL rabbit after balloon denudation of the abdominal aorta with the external phased-array MR coil.
Modification or stabilization of vulnerable plaques in the coronary arteries by strengthening the fibrous cap and decreasing the lipid core has been proposed as one of the mechanisms responsible for the observed beneficial clinical effect of lipid-lowering therapies.2 Thus, in the future, one might be able to sequentially image and monitor such compositional atherosclerotic plaque changes with MRI.3 The improvements in resolution afforded by an IV MR coil may be crucial for allowing accurate quantification of atherosclerotic plaque components in human coronary arteries.
We did not perform a formal analysis of the vessel wall parameters with the surface coil for the purposes of comparison with the IV coil parameters. Our purpose in this study was to establish the feasibility of this new IV MR coil for high-resolution imaging of the arterial wall. Thus, analyses with histopathology and the IV coil measurements were performed to confirm this. We4,5⇓ and others12,13⇓ have already published studies showing the ability of surface coil imaging with MR to accurately quantify atherosclerosis in the rabbit model. However, by establishing the feasibility of this nonobstructive IV coil to accurately obtain high-resolution images of the arterial wall, we establish the potential for its use in the human coronary system.
We noted the ability of the IV coil to identify both high- and low-signal structures within the atherosclerotic artery wall, consistent with fibrotic and lipidic plaque components, respectively. However, the purpose of this study was not to show that MRI could accurately identify and quantify components of complex atherosclerosis, because many groups have already published this finding.11,15,21⇓⇓ However, we wished to show the quality and resolution of the images obtainable, without significant motion degradation, with this novel IV MR coil.
IV MR coils have been reported for vascular imaging and guiding interventional procedures.9,22–25⇓⇓⇓⇓ Some of these devices required the total occlusion of the coronary arteries. The availability of a nonobstructive design permits longer acquisition times, which would be required for fast spin-echo imaging of arterial wall structures. However, flow characteristics and motion in the absence of a centering balloon requires additional study with this IV MR coil.
In summary, we are reporting the feasibility of a nonobstructive IV MR coil with improved resolution over surface phased-array coils for the in vivo imaging of atherosclerotic lesions. Continued improvements in MRI techniques and additional studies are required, however, to confirm the ability of this IV MR coil to be used for this purpose in human coronary arteries.
This work was supported by grants from the National Heart Foundation of Australia (SA Branch) (S. Worthley), The French Federation of Cardiology (Dr Helft), and the National Institutes of Health (SCOR NIH P50 HL54469) to Dr Fuster, Dr Fallon, and J. Badimon.
Received December 3, 2002; revision accepted December 4, 2003.
- ↵Libby P, Aikawa M. New insights into plaque stabilisation by lipid lowering. Drugs. 1998; 56: 9–13.
- ↵Worthley SG, Helft G, Fuster V, Zaman AG, Fayad ZA, Fallon JT, Badimon JJ. Serial in vivo MRI documents arterial remodeling in experimental atherosclerosis. Circulation. 2000; 101: 586–589.
- ↵Worthley SG, Helft G, Fuster V, Fayad ZA, Rodriguez OJ, Zaman AG, Fallon JT, Badimon JJ. Noninvasive in vivo magnetic resonance imaging of experimental coronary artery lesions in a porcine model. Circulation. 2000; 101: 2956–2961.
- ↵Fayad ZA, Fuster V, Fallon JT, Jayasundera T, Worthley SG, Helft G, Aguinaldo JG, Badimon JJ, Sharma SK. Noninvasive in vivo human coronary artery lumen and wall imaging using black-blood magnetic resonance imaging. Circulation. 2000; 102: 506–510.
- ↵Botnar RM, Stuber M, Kissinger KV, Kim WY, Spuentrup E, Manning WJ. Noninvasive coronary vessel wall and plaque imaging with magnetic resonance imaging. Circulation. 2000; 102: 2582–2587.
- ↵Zimmermann-Paul GG, Quick HH, Vogt P, von Schulthess GK, Kling D, Debatin JF. High-resolution intravascular magnetic resonance imaging: monitoring of plaque formation in heritable hyperlipidemic rabbits. Circulation. 1999; 99: 1054–1061.
- ↵Rogers WJ, Prichard JW, Hu YL, Olson PR, Benckart DH, Kramer CM, Vido DA, Reichek N. Characterization of signal properties in atherosclerotic plaque components by intravascular MRI. Arterioscler Thromb Vasc Biol. 2000; 20: 1824–1830.
- ↵McConnell MV, Aikawa M, Maier SE, Ganz P, Libby P, Lee RT. MRI of rabbit atherosclerosis in response to dietary cholesterol lowering. Arterioscler Thromb Vasc Biol. 1999; 19: 1956–1959.
- ↵Worthley SG, Helft G, Fuster V, Fayad ZA, Fallon JT, Osende JI, Roque M, Shinnar M, Zaman AG, Rodriguez OJ, Verhallen P, Badimon JJ. High resolution ex vivo magnetic resonance imaging of in situ coronary and aortic atherosclerotic plaque in a porcine model. Atherosclerosis. 2000; 150: 321–329.
- ↵Toussaint JF, LaMuraglia GM, Southern JF, Fuster V, Kantor HL. Magnetic resonance images lipid, fibrous, calcified, hemorrhagic, and thrombotic components of human atherosclerosis in vivo. Circulation. 1996; 94: 932–938.
- ↵Yuan C, Beach KW, Smith LH, Jr, Hatsukami TS. Measurement of atherosclerotic carotid plaque size in vivo using high resolution magnetic resonance imaging. Circulation. 1998; 98: 2666–2671.
- ↵Hatsukami TS, Ross R, Polissar NL, Yuan C. Visualization of fibrous cap thickness and rupture in human atherosclerotic carotid plaque in vivo with high-resolution magnetic resonance imaging. Circulation. 2000; 102: 959–964.
- ↵Fayad ZA, Nahar T, Fallon JT, Goldman M, Aguinaldo JG, Badimon JJ, Shinnar M, Chesebro JH, Fuster V. In vivo magnetic resonance evaluation of atherosclerotic plaques in the human thoracic aorta: a comparison with transesophageal echocardiography. Circulation. 2000; 101: 2503–2509.
- ↵Yuan C, Mitsumori LM, Ferguson MS, Polissar NL, Echelard D, Ortiz G, Small R, Davies JW, Kerwin WS, Hatsukami TS. In vivo accuracy of multispectral magnetic resonance imaging for identifying lipid-rich necrotic cores and intraplaque hemorrhage in advanced human carotid plaques. Circulation. 2001; 104: 2051–2056.