Noninvasive imaging of atherosclerotic vessels by MRI

Roberto Corti
MD

Isabella Sudano
MD
Cardiology Department
University Hospital Zurich, Switzerland

Atherothrombosis is a systemic arterial disease originally involving the intima of large- and ­medium-sized systemic arteries, including the carotid, aorta, coronary and peripheral arteries.

Since the composition of “high-risk plaques” varies depending on the arterial region and because there is striking heterogeneity in the composition of human atherothrombotic plaques, even within the same individual, reliable noninvasive imaging modalities able to detect atherothrombotic disease in the various stages and regions and characterise the composition of the plaques are clinically desirable.(1,2) The assessment of atherosclerotic plaques by imaging techniques is essential for the in-vivo identification of high-risk plaques.(3–5) Several invasive and noninvasive imaging techniques are available to assess atherosclerotic disease. Most of the standard techniques identify luminal diameter or stenosis, wall thickness and plaque volume. The angiogram – performed by X-ray, computed tomography (CT) or magnetic resonance imaging (MRI) – reflects luminal diameter and provides, with excellent resolution, a measure of stenosis or information of the luminal surface of protruding atherothrombotic disease.(6,7) High-resolution MRI has emerged as the potential leading noninvasive in-vivo imaging modality for atherosclerotic plaque characterisation. MRI is noninvasive, does not involve ­ionising radiation and can provide high-resolution images of multiple vascular territories. Improvements in MR techniques(2,8) (eg, faster imaging and detection coils), conducive to high resolution and contrast imaging, have permitted the study of the different plaque components in the human carotid,(1,8–12) aortic,(1,12,13) peripheral(14,15) and coronary arterial disease.(16–18)

Magnetic resonance for plaque imaging
The assessment of both atherosclerotic plaque anatomy and composition has been validated in experimental models and in humans. Validation studies in experimental models of atherosclerosis have been performed in rabbits,(19,20) in pigs, using clinically available MR, and in mice,(21) using high-field-strength MR. Worthley et al were the first to validate the MR technique for coronary plaque imaging ex vivo(16) and in vivo(22) in a clinical MR scan using a porcine model.

In humans, validation was initially performed ex vivo using histological specimens(23,24) and subsequently in vivo in patients scheduled for carotid endarterectomy(8–11,25–27) or before surgical grafting of abdominal aortic aneurysm,(28) in which in-vivo MR images were compared with histology. Fayad et al were the first to demonstrate the feasibility of coronary plaque imaging in human in vivo.(17) This technique was subsequently improved by Botnar et al,(29) who were able to perform high-resolution coronary plaque imaging during free breathing.

Measurement of vessel wall dimensions
Yuan et al reported the accuracy of vessel wall measurement in vivo by MRI,(11) showing for the first time that this imaging technique has potential application in monitoring lesion size in studies examining plaque progression and/or regression. Corti et al(12) reported similar in-vivo data on the reproducibility of the vessel wall area measurement.

Fayad et al(13) used MRI to assess thickness, extent and composition of atherosclerosis in the ­thoracic aorta and show that the results obtained with noninvasive MR correlate well with transoesophageal echocardiography imaging.

Evaluation of plaque composition
Skinner et al showed for the first time that progression of disease, resulting in increase in lesion mass, decrease in arterial lumen and intralesion complications, can be detected by advanced MR techniques in the rabbit abdominal aorta.(20) Images acquired in vivo correlated with the fine structure of the lesions of atherosclerosis, including the fibrous cap, necrotic core and lesion fissures, as verified by gross examination, dissection microscopy and histology.

[[HHE07_fig1_Ca13]]

Several groups followed this pioneer work and further implemented the technique, allowing ­characterisation of the major plaque components. Toussaint et al(23) found that T2-weighted imaging could discriminate between the collagenous cap and lipid core of atherosclerotic lesions. Fayad et al(13) showed that MR accurately determined wall area in comparison with histopathology in apolipoprotein E knockout mice and showed excellent agreement in grading of lesion shape and type.(21)

Helft et al(30) used the rabbit model to study the ability of noninvasive in-vivo MRI to quantify the different components within atherosclerotic plaque (such as calcification, fibrocellural tissue, fibrous tissue with extracellular lipid, lipid cores and thrombus). A high level of agreement between in-vivo MRI and histological examination of tissue for the identification of plaque components and fibrous cap ­characteristics has been demonstated in humans.(23–25,31–33) Fayad et al(13) showed that MRI allows ­assessement of plaque size and composition in the thoracic aorta, while Shinnar et al(34) demonstrated it for the carotid plaque constituents.(13) In human carotid artery, MRI allows imaging and ­characterisation of ­normal and pathological arterial walls,(25) the quantification of plaque size(11) and the detection of fibrous cap integrity.(35) In human peripheral arteries we showed that cross-sectional MRI performed 24 hours after percutaneous transluminal angioplasty at the level of the arterial occlusion revealed severe ­disruption and splitting of the atherosclerotic plaque, resulting in an irregular-shaped lumen (see Figure 2).(14) More recently, a new self-contained intravascular MR probe was used to assess plaque composition in the coronary artery.(36)

[[HHE07_fig2_Ca13]]

Interventional trials
Several investigators have recently used serial noninvasive MRI to assess in vivo the effects of interventional strategies (such as dietary interventions, systemic medical therapy or percutaneous balloon treatment) on animal models of arteriosclerosis and in humans.

In-vivo monitoring of therapy with MRI in experimental arteriosclerosis Aortic plaque progression was seen in rabbits maintained on a high-cholesterol diet, whereas plaque regression followed after ­resuming a low-cholesterol diet.(37)

Corti et al(38) recently reported plaque regression and features of plaque stabilisation by a ­combination of statin and a new selective peroxisome proliferator-activated receptor γ (PPARγ)-agonist in the atherosclerotic rabbit model (see Figure 3). Viles-Gonzalez et al(39) showed that a novel antithrombotic therapy, by inhibition of the thromboxane A(2) receptor, caused a regression of advanced atherosclerotic plaques in rabbits (see Figure 4).

[[HHE07_fig3_Ca14]]

[[HHE07_fig4_Ca14]]

In-vivo monitoring of therapy with MRI in humans: effects of lipid-lowering by statin Corti et al(12) have shown that in-vivo MR can be used to measure the effect of lipid-lowering therapy in asymptomatic untreated hypercholesterolaemic patients with carotid and aortic atherosclerosis. Atherosclerotic plaques were visualised and measured with MR at different time points after initiation of lipid-lowering therapy. Significant regression of atherosclerotic lesions, a decrease in the vessel wall area but no change in the lumen area were observed after 12-month therapy. A longer follow-up showed a continued ­reduction in arterial wall area and even a small, but significant, increase in the arterial lumen at 24 months (see Figure 5).(40) In a more recent study using a similar design in patients with coronary artery disease, ­significant regression was already seen at six months.(41) Using the same study design, Yonemura et al(42) showed that a marked LDL-cholesterol reduction was associated with regression of thoracic aortic plaques but only retardation of plaque progression in abdominal aorta. In another remarkable study, Zhao et al(43) examined carotid plaque composition quantitatively by MRI, showing that treatment with lovastatin, niacin and colestipol may significantly reduce the lipid core from the lesions. Moreover, high-­resolution MRI may identify the modifications of the atherosclerotic plaque and of the arterial wall following ­percoutanous angioplasty and endovascular brachytherapy in human peripheral artery.(14,15,44)

[[HHE07_fig5_Ca15]]

Issues and challenges
Despite the rapid improvements of noninvasive techniques such as MRI for the in-vivo evaluation of atherothrombosis, their application at clinical level is still hampered by certain issues that deserve further intensive research. For instance, studies are needed to elucidate the inter- and intra-individual variability in the distribution of atherosclerotic plaques and the ability of noncoronary atherosclerotic plaques imaging to predict cardiovascular events. In fact, most of the data presently available derives from angiographic or autopsy studies. The availability of robust noninvasive techniques to image atherosclerotic vessels will allow comprehensive studies of atherosclerosis distribution and the correlation between plaque burden and cardiac events, and cerebrovascular and peripheral events, respectively. In addition, we are still missing adequate data on the spontaneous course of atherosclerosis. Against intuitive thinking, progression of atherothrombosis is most likely to be described by nonlinear growth. At present the mechanisms of progression remain very speculative.

In addition, at present it is still not known whether spontaneous regression of atherosclerotic lesions is also possible in adults. In fact, in young men regression of early atherosclerotic lesions (such as fatty streaks) appears to be possible. An additional crucial issue is whether the detection of high-risk plaques will enable the prediction of cardiovascular events. Despite very rapid technical evolution of all noninvasive imaging modalities, their application in the coronary circulation remains limited and mainly hampered by the spatial and temporal resolution.

Conclusion
The assessment of atherosclerotic burden by imaging techniques appears to be an essential tool for the identification of high-risk plaques and the risk stratification of individual patients. Although not yet available for routine use, in-vivo, high-resolution, multicontrast MRI is the most promising method of noninvasively imaging plaques and characterising the main plaque components. Noninvasive MRI also allows serial evaluation of the progression and regression of atherosclerosis over time. Therefore, this technology is particularly appealing to test the effects of novel antiatherosclerotic drugs using plaque morphology as a surrogate endpoint before investing in hard-endpoint trials. MRI technology is ­evolving rapidly and will open up new areas for the diagnosis, prevention and treatment of atherosclerosis in all arterial locations.

References

1. Ann NY Acad Sci 2001;947:181-95;195-8.
2. Circ Res 2001;89(4):305-16.
3. Eur Heart J 2002;4,4:A31-41.
4. Am J Cardiol 2003;91(3A):17-26A.
5. Curr Opin Cardiol 2002;17(6):616-25.
6. Lancet 2001;357(9262):1086-91.
7. Circulation 2002;105(3):328-33.
8. Radiol 2001;221(2):285-99.
9. J Magn Reson Imaging 2002;15(1):62-7.
10. Circulation 2001;104(17):2051-6.
11. Circulation 1998;98(24):2666-71.
12. Circulation 2001;104(3):249-52.
13. Circulation 2000;101(21):2503-9.
14. Circulation 2002;106(12):1570-1.
15. Heart 2000;83(2):188-91.
16. Atherosclerosis 2000;150(2):321-9.
17. Circulation 2000;102(5):506-10.
18. Circulation 1999;99(24):3139-48.
19. J Am Coll Cardiol 2001;37:1149-54.
20. Nature Medicine 1995;1(1):69-73.
21. Circulation 1998;98(15):1541-7.
22. Circulation 2000;101(25):2956-61.
23. Arterioscler Thromb Vasc Biol 1995;15(10):1533-42.
24. Arterioscler Thromb 1994;14(12):1951-7.
25. Circulation 1996;94(5):932-8.
26. Proc Int Soc Magn Reson Med 1999;1:84.
27. Circulation 2002;105(2):181-5.
28. Circulation 2004;109(8):1016-21.
29. Circulation 2000;102(21):2582-7.
30. J Am Coll Cardiol 2001;37(4):1149-54.
31. Invest Radiol 1986;21(5):411-5.
32. Invest Radiol 1989;24(1):52-60.
33. Arterioscler Thromb Vasc Biol 1997;17(3):542-6.
34. Arterioscler Thromb Vasc Biol 1999;19(11):2756-61.
35. Circulation 2000;102(9):959-64.
36. J Am Coll Cardiol 2006;47;C48-56.
37. Arterioscler Thromb Vasc Biol 1999;19(8):1956-9.
38. J Am Coll Cardiol 2004;43(3):464-73.
39. Eur Heart J 2005;26:1557-61.
40. Circulation 2002;106(23):2884-7.
41. Circulation 2004;110(16):2336-41.
42. J Am Coll Cardiol 2005;45(5):733-42.
43. Arterioscler Thromb Vasc Biol 2001;21(10):1623-9.
44. Circulation 2004;110(9):1156-61.