OUP user menu

Selective coronary artery plaque visualization and differentiation by contrast-enhanced inversion prepared MRI

David Maintz, Murat Ozgun, Andreas Hoffmeier, Roman Fischbach, Won Yong Kim, Matthias Stuber, Warren J. Manning, Walter Heindel, René M. Botnar
DOI: http://dx.doi.org/10.1093/eurheartj/ehl102 1732-1736 First published online: 20 June 2006


Aims We sought to evaluate the utility of contrast-enhanced coronary magnetic resonance imaging (CE-MRI) for selective visualization and non-invasive differentiation of atherosclerotic coronary plaque in humans.

Methods and results Nine patients with coronary artery disease (CAD) as confirmed by X-ray angiography and multidetector computed tomography (MDCT) were studied by T1-weighted black blood inversion recovery coronary MRI before (N-IR) and after administration of Gd-DTPA (CE-IR). Plaques were categorized as calcified, non-calcified, and mixed based on their Hounsfield number derived from MDCT. With MDCT, a total of 29 plaques were identified, including calcified (n=6), non-calcified (n=6), and mixed calcified/non-calcified (n=17). On N-IR MRI, 26 plaques (90%) were dark, whereas three plaques (two non-calcified and one mixed) appeared bright. On CE-MRI, 13/29 (45%) plaques, 11 of which were mixed, one non-calcified, and one calcified showed contrast uptake. All others remained dark.

Conclusion In this preliminary study, we demonstrate the potential utility of CE-IR MRI for selective plaque visualization and differentiation of plaque types. The observed contrast uptake may be associated with endothelial dysfunction, neovascularization, inflammation, and/or fibrosis.

  • Magnetic resonance imaging
  • Coronary plaque
  • Contrast enhancement
  • Plaque characterization


Atherosclerosis is a chronic systemic inflammatory disease of the vessel wall affecting large, medium, and small arteries such as the aorta, the carotid, and coronary arteries. During the past decade, progression and complication of coronary artery atherosclerosis has been intensively studied and stages of plaque progression have been defined.1 Evidence has been found that most acute coronary syndromes are caused by rupture of atherosclerotic plaques of AHA stages 3 to 4 without underlying high grade lumen stenosis.2 Invasive X-ray coronary angiography (XCA) is the current gold standard for visualization of the coronary artery lumen, but does not visualize and grossly underestimates the underlying plaque and is therefore not suitable for early disease detection or characterization of plaque components. Potential for non-invasive plaque characterization has recently been assigned to multidetector computed tomography (MDCT) and cardiovascular magnetic resonance (CMR). MDCT can distinguish between calcified plaques, mixed plaques with calcified and non-calcified components, and non-calcified plaques.3 MRI has the advantage of a superior soft tissue contrast. In vivo coronary vessel wall visualization has successfully been demonstrated using CMR and coronary artery remodelling in patients with X-ray confirmed coronary artery disease (CAD) has been shown.46

Contrast-enhanced MRI (CE-MRI) has been demonstrated to improve the contrast-to-noise (CNR) ratio between the vessel lumen and the vessel wall, but previous studies using gadolinium (Gd)-based or iron-based contrast agents have been restricted to larger arteries79 or animal models.1012 We sought to examine the potential utility of CE-MRI for plaque visualization and characterization using a T1-weighted CMR technique with an inherent high contrast between contrast enhancing and non-enhancing tissue.


Study population and design

The study population included nine prospectively enrolled patients (6 male, 3 female, age 44–72 years, mean 61) with known CAD. All subjects had previously undergone XCA and coronary 16-slice MDCT angiography. The time interval between XCA and CMR was 1–29 days (mean 15 days) and between MDCT and CMR was 1 day. All patients initially assessed for inclusion deemed eligible for the study. The baseline and main clinical characteristics of study patients are summarized in Table 1.

View this table:
Table 1

Baseline and main clinical characteristics of the patients

Age (years)61±9.5
Cardiovascular risk factors9/9
 Diabetes type I1
 Diabetes type II1
 Positive family history2
Prior myocardial infarction3
Stable angina pectoris6
  • BMI, body mass index. Age and BMI given as mean±SD.

The study complies with the Declaration of Helsinki and the study protocol was approved by the locally appointed Ethics Committee and all patients gave written informed consent to participate.

Multidetector spiral computed tomography

MDCT imaging was performed on a 16-detector row scanner (Sensation 16, Siemens Medical Systems, Forchheim, Germany) applying a detector collimation of 16×0.75 mm with a table feed of 3.4 mm/rotation at a rotation time of 420 ms. The tube current was 370 mAs and the tube voltage 120 kV. Patients with a heart rate above 65 bpm. received oral medication of 80 mg propranolol 60 min before the examination. Arterial contrast was achieved with intravenous application of 100 mL of a non-ionic contrast medium (Ultravist 300, Schering, Berlin, Germany) injected at a flow rate of 4 mL/s. Transaxial images (slice thickness 1.0 mm, increment 0.5 mm) were reconstructed using an ECG-gated half-scan reconstruction algorithm (temporal resolution 210 ms) and a medium-sharp kernel (B30f). The position of the reconstruction window within the cardiac cycle was individually optimized to minimize motion artifacts.13

CMR imaging

CMR imaging was performed on a 1.5T MR system (Intera, Philips Medical Systems, Best, The Netherlands) using a commercial five element cardiac synergy coil. After a survey scan to localize the heart and diaphragm, a multi-heart phase SSFP cine sequence (TR, 2.9 ms; TE, 1.5 ms; flip angle, 60°; 50 heart phases; matrix, 144×144; SENSE factor 2) was obtained to assess the interval of minimal right coronary artery (RCA) motion for the determination of the trigger delay of the subsequent coronary CMR lumen and vessel wall scans.

Coronary bright-blood MRA was performed with a previously described navigator-gated free-breathing and cardiac-triggered T2-prepared three-dimensional steady-state free-precession sequence allowing visualization of the anatomy of the coronary artery lumen.14

The following coronary plaque scan was a navigator-gated free-breathing and cardiac-triggered T1-weighted inversion-recovery and fat-suppressed 3D black-blood gradient-echo sequence. Parameters of the sequence were: TR, 6.1 ms; TE, 1.9 ms; flip angle, 30°. Spatial resolution of both sequences was 1.0×1.0×3.0 mm. Scans of the proximal left and RCA system were performed in the transverse plane to allow for slice-to-slice comparison with MDCT data. The number of slices was limited to 20 because of time constraints.

After intravenous application of 0.3 mmol/kg body weight of a Gd-based contrast medium (Magnevist, Schering, Berlin, Germany) and a predefined delay time between contrast administration and imaging of 3 h outside the scanner, the above described scan protocol was repeated.

The inversion time (TI) of the inversion recovery sequences was individually assessed according to the T1 of blood using a TI scout (Look Locker sequence, 15). Typically, the inversion time before contrast administration was 450 ms and after contrast application 250 ms.15,16

Data analyses

Two investigators evaluated the MDCT data sets in consensus using axial and multiplanar reformatted images. For every coronary artery segment, identified via side branches, the investigators decided whether calcified plaque, non-calcified plaque, both or neither was present in MDCT.3

Original MDCT- and CMR-source images were evaluated by two experienced readers in consensus on a segment-per-segment basis. Both readers were blinded to the catheter angiography results and clinical data.

Contours were drawn around areas of contrast uptake on post-contrast MR images and copied to the corresponding pre-contrast images. MR contrast enhancement was defined as a signal-to-noise ratio (SNR defined as signal of plaque divided by SD of noise) increase of at least 50% and a CNR (defined as signal of plaque minus signal of blood divided by SD of noise) increase of at least 100% with respect to the native N-IR scan. Blood signal was measured at the root of the aorta in pre- and post-contrast images. Noise was defined as the SD measured in air anterior to the patients chest wall.


All nine MDCT and CMR examinations were completed without complication. While the scan volumes of MDCT covered the whole heart, the coronary MRA and the CMR plaque scans were limited in volume covering a total of 55 proximal and mid-coronary artery segments available for further analysis [in all patients the proximal RCA, the left main (LM) coronary artery, the proximal left anterior descending coronary artery (LAD), and proximal circumflex coronary artery (Cx) were covered]. The mid-LAD was included in the scan volumes of six patients, the mid-Cx in four, and the mid-RCA in four patients. In addition, the first diagonal branch (D1) was included in three and the first marginal branch (M1) in two patients.

Within these segments, 29 plaques were identified using MDCT and classified as calcified (n=6), non-calcified (n=6), or mixed calcified/non-calcified (n=17; Table 2). The plaques were located and distributed among the following segments: proximal RCA, 5; mid-RCA, 3; LM, 3; proximal LAD, 6; mid-LAD, 4; proximal Cx, 3; mid-Cx, 3; D1, 1; M1, 1.

View this table:
Table 2

Differentiation of different plaque types with regard to MDCT morphology and CMR signal intensity before and 3 h after contrast administration

CT morphologyNo.DarkBrightEnhancement

The inversion recovery CMR sequence facilitated suppression of both blood and surrounding tissues, thereby enabling selective plaque visualization. After contrast administration, 13 plaques demonstrated enhancement (Figures 1 and 2), including 11 (65%) of 17 mixed calcified/non-calcified but only 1 (17%) of calcified and 1 (17%) of non-calcified plaques.

Figure 1

Detection of coronary artery plaque enhancement. High-grade ostial stenosis of the RCA as demonstrated by catheter angiography (A) and MR-angiography (D). MDCT coronary angiography (B) shows that the underlying plaque is severely calcified, but also has non-calcified components (C). Inversion recovery CMR-images before (E) and after (F) contrast media application demonstrate strong focal enhancement of the plaque and diffuse enhancement of the RCA wall.

Figure 2

Enhancement of coronary artery plaque. Proximal plaque of the RCA with calcified and non-calcified components as demonstrated by MDCT (A) causing medium- to high-grade lumen stenosis in MR-angiography (B). Inversion recovery CMR-images before (C) and after (D) contrast media application demonstrate focal enhancement of the RCA plaque (straight arrow). Please also note strong enhancement of aortic plaque (dotted arrow).

Quantitative signal measurements

Of the 29 plaques, 26 (90%) were hypointense on N-IR images with an SNR of 4.9±1.5. Three plaques had a hyperintense appearance on N-IR imaging corresponding to an SNR >10 (14.8±1.6).

After contrast agent administration, 13 of 29 (45%) plaques showed significant enhancement corresponding to an SNR and CNR increase of 233±149 and 445±292%. Average SNR and CNR of those plaques on N-IR were 4.6±1.1, 2.1±1.8, 15.4±7.6, and 12.2±7.1, respectively.

The remaining 16 plaques did not exhibit significant contrast uptake (SNR 33±33% and CNR 60±51%). The three plaques that appeared bright at baseline were among those 16 plaques without significant enhancement. Figure 3 shows one example for each of the plaque types according to their signal intensities before and after contrast agent application.

Figure 3

Examples of different plaque types. From left to right: catheter angiography, MDCT coronary angiography, coronary MRI, CMR before contrast-media application, and CMR 3 h after contrast-media application. (A) LAD coronary artery plaque with mixed components in MDCT appears dark before and after contrast application (type 1). (B) Marginal branch plaque (non-calcified) appearing bright before and after contrast application (type 2). (C) RCA plaque with low signal before and bright signal after contrast enhancement (type 3).


This study of patients with CAD yielded two important findings:1 The use of black blood CE-IR coronary MRI for the detection of selective contrast uptake in non-calcified and mixed coronary plaque in patients with CAD has been demonstrated.2 The different signal intensities of plaques in N-IR scans in combination with different signal intensities after contrast application facilitates differentiation of three different plaque types (Figure 3): (i) plaques that are hypointense on N-IR and CE-IR (type 1); (ii) plaques that are hyperintense both on N-IR and on CE-IR (type 2); and (iii) plaques that are hypointense on N-IR and hyperintense on CE-IR (type 3).

CMR technique

CMR has been accredited for high potential to non-invasively image the coronary artery vessel wall.46 While these previous studies used techniques for unselective visualization of the entire vessel wall (in both healthy and diseased states), our approach facilitates visualization of coronary artery plaque morphology. As the imaging task using the proposed method is reduced to the assessment of the presence or absence of contrast uptake, requirements on spatial resolution are less stringent. Furthermore, image interpretation is reduced to the detection of hot spots similar to radionuclide methods.

Enhancement phenomenon

To our knowledge, this is the first report of in vivo contrast enhancement of coronary artery plaques in human coronary arteries using a Gd-based MR contrast medium. In previous studies, gadolinium uptake has been observed in patients with advanced carotid atherosclerosis using a T1-weighted fast spin-echo imaging technique.8,9 In contrast to the here proposed IR-prepared segmented gradient-echo sequence, contrast uptake is more difficult to appreciate because of the reduced suppression of surrounding tissues. Other investigators have used contrast agents based on iron oxides that lead to a negative contrast on T2-weighted images after phagocytosis by plaque-mediated macrophages. Similar to the Gd-based studies, imaging was performed in human carotid arteries or in animal models.10,17

In our protocol, MR imaging was performed immediately before and 3 h after i.v. injection of contrast medium application. The ‘late enhancement’ approach ensures that the contrast medium had already been renally filtered and that physiologic uptake of contrast medium in healthy tissues has washed out. Previous studies have shown that both inflammation1820 and fibrosis21 are associated with contrast uptake and may lead to the same visual appearance. This delayed enhancement effect can be explained by an increased extracellular space and by a reduced washout kinetics. These studies also demonstrate a relationship between inflammation and fibrosis, whereby fibrosis was found in patients with chronic or recurrent inflammation.19 Differentiation between these two entities may be possible based on differences in the pharmacokinetics. Thus we hypothesize that the observed contrast uptake may be associated with inflammatory or fibrous coronary plaques and thus may reflect acute as well as chronic vascular inflammation. Inflammation is a known process in the development and progression of atherosclerotic plaques and is suspected to play a key role in plaque rupture, suggesting that plaque enhancement may identify potentially vulnerable plaques. Therefore, plaque enhancement might indicate potentially vulnerable plaques. These assumptions remain to be investigated in subsequent studies and histological correlation is required for verification.

Plaque types

For our study, MDCT was used as the reference standard for the detection of coronary plaques. This method enables classification of plaques as calcified, non-calcified, and mixed calcified and non-calcified. Using our new CE-MR protocol, we found that one group of non-calcified and mixed plaques appear bright on CE-MRI while a substantial minority do not display enhancement. Although our study lacks histological correlation of our findings, we hypothesize on the underlying composition of the plaques that lead to the different MR appearance. Tissue that appears bright in T1-weighting must have a very short T1-relaxation time. One tissue that might fulfil this criterion is haemorrhage containing methemoglobin. Similarly very high signal of certain plaque types on strongly T1-weighted images have been reported for the carotid arteries22 with a histological correlation with methemoglobin in intraplaque haemorrhage. As intraplaque haemorrhage is a criterion of complicated plaque (type VI), our technique may enable non-invasive detection of advanced plaque stages and/or recent plaque rupture. Intraluminal thrombus, with fewer trapped red blood cells, may have different signal characteristics when compared with a haemoglobin-rich intraplaque haemorrhage.23


This study demonstrates selective coronary plaque visualization by CE-MRI that allows for differentiation between MDCT-characterized plaque entities. Further investigations are now warranted to investigate these findings in larger patient studies to assess the methods sensitivity and specificity for the detection of plaque and histological correlation is needed for clarification of the different plaque signal intensities.

Conflict of interest: none declared.


View Abstract