Background
Cardiovascular Magnetic resonance (CMR) is a non-invasive imaging modality that has enabled the assessment of both the morphological [
1] and compositional characteristics of the carotid arterial wall. Via histological validation, multi-sequence high-resolution carotid CMR has been proven to characterize fibrous tissue, lipid-rich necrotic core, calcification, and hemorrhage in the human atherosclerotic plaque
in vivo [
2‐
4]. Moreover, the high reproducibility [
5‐
7] of carotid CMR has made possible the effective quantitative assessment of plaque evolution [
8] and the identification of compositional features associated with accelerated progression [
9].
Studies with carotid CMR, however, have been limited to the evaluation of subjects with established carotid disease [
10,
11]. Although these and other carotid studies [
12] have proven invaluable in the assessment of local plaque phenomena, the relationship between CMR identified carotid disease and atherosclerotic disease in other arterial beds has not previously been reported. Associations between the carotid artery and other vascular beds have been described in the ultrasound literature [
13]. In particular, multiple investigations have successfully employed B-mode ultrasound to identify differences in carotid intima-media thickness (IMT) between individuals with and without coronary artery disease (CAD) [
14,
15].
We designed the C arotid A therosclerosis (M RI) P rogression S tudy (CAMPS) to evaluate the presence and progression of carotid atherosclerosis measured by CMR over 2 years in patients with and without obstructive CAD. CAMPS is an observational, prospective investigation and participants were not selected based on carotid status. In the study described herein, we evaluated baseline data from this cohort and tested the hypothesis that individuals with and without obstructive CAD differ in their carotid arterial wall morphology and composition as identified by high spatial resolution carotid CMR. We further sought to determine if these differences were unique to particular segments of the carotid artery and to identify any gender specific findings.
Discussion
This case-control study identified several new differences in carotid arterial structure and composition between patients with and without obstructive CAD. Males (but not females) with obstructive CAD had significantly smaller carotid arteries with thicker walls and an increased prevalence of lipid-rich necrotic core. These findings were limited to the carotid bulb and ICA. In contrast to these gender-specific differences, calcification was related to CAD status in both males and females.
The association between intima-media thickness of the carotid wall and CAD status has been well described in the ultrasound literature [
13‐
15,
20] and parallels our findings for CAD-related disparity in carotid wall thickness. In addition, however, we describe differences in area-based metrics (lumen area, total vessel area and NWI) between CAD cases and controls. The direction of these area-based differences was unexpected. From the study of coronary arteries, Glagov et al [
21] proposed that during the expansion of early atherosclerotic lesions, preservation of luminal area occurred secondary to outward remodeling. The
smaller total vessel area and lumen area that we detected in the CAD case group, while consistent with reported B-mode ultrasound measurements of ICA diameters from a similar cohort [
22], appears inconsistent with the outward remodeling theory. This divergence is unlikely a result of the CAD cases having inward remodeling as a result of advanced carotid atherosclerotic disease since the wall area between groups was very similar. Without serial data to monitor the evolution of change, we can only speculate as to the etiology of the observed differences in arterial morphology. One possibility is that there may be inherent differences in arterial structure between the carotid arteries evaluated here and the coronary arteries that provided data for the remodeling hypothesis. Alternatively, the smaller carotid arteries of CAD cases may reflect impaired nitric oxide function manifested at a very early phase of atherosclerosis [
23]. Lastly, individuals, particularly males, with smaller arteries at baseline (i.e. before they develop CAD or carotid disease) may be more prone to develop atherosclerotic disease. In this latter hypothesis, outward remodeling as predicted by Glagov may have resulted in the thickened arterial walls that were observed in the obstructive CAD cases.
A natural extension of the area-based metrics is NWI, which normalizes plaque burden to arterial size and is similar to the percent atheroma volume measurement proposed by Nissen et al for quantification of coronary atherosclerosis [
24]. In previous CMR studies, NWI has been shown to be the most reproducible measure of plaque burden [
5]. In this investigation, NWI provided the strongest discriminator of morphological differences between CAD cases and controls. Although there were differences in NWI at all segments between groups in males, the most robust differences in NWI and in the other metrics occurred in the distal bulb and ICA. Our findings suggest that early carotid disease may be best identified in the distal bulb and ICA. Furthermore, NWI may be the most sensitive CMR metric for detection of early atherosclerotic disease in any segment.
Except for NWI in the ICA, significant differences in arterial structure between groups were limited to males. Gender-based disparity in IMT is well recognized, but a gender-specific distinction in carotid structure between CAD cases and controls has not been previously reported. Crouse et al has previously documented a relationship between male gender and IMT progression rates in CAD cases [
25]. These new observations indicate that carotid wall structure for this particular demographic may be more strongly associated with CAD in males than females.
Beyond structural measurements, an additional advantage of CMR is its ability to characterize arterial wall composition. In parallel with the differences between CAD cases and controls in arterial structure, presence of a lipid-rich necrotic core, particularly in the distal bifurcation, was associated with the presence of obstructive CAD in males. Integrated backscatter analysis in B-mode ultrasound has previously been used to demonstrate a difference in extent of carotid atheromatous tissue between patients with a history of myocardial infarction and low-risk controls [
26]. These complimentary findings provide compelling evidence that the lipid-rich necrotic core is not only the basis of local carotid plaque instability [
27], but that individuals with CAD may have an increased likelihood of developing plaques that have more pronounced lipid-rich necrotic cores.
Finally, calcification was a gender-independent marker and the most robust compositional indicator of obstructive CAD status. The association between calcification and cardiovascular disease has previously been demonstrated in the coronary arteries. Coronary artery calcification as measured by cardiac computed tomography has been closely correlated with severity of coronary artery disease at autopsy [
28,
29] and has been shown to predict cardiovascular events in asymptomatic individuals [
30]. Our findings in the carotid artery are consistent with cardiac computed tomography data and indicate that the presence of calcification in either arterial bed may be indicative of CAD.
Limitations
The strengths and limitations of using angiographically defined case-control groups have been previously described [
31,
32]. The control group may not accurately reflect an at-large population because of 1) the underlying reason for their referral to angiography, and 2) their proven absence of obstructive coronary disease. In addition, individuals with CAD generally receive intensive medical treatment by their primary care physician. Consequently, these results should not be extrapolated to the general population. Individuals with non-obstructive CAD (<50% stenosis) were not enrolled and did not receive a carotid CMR in this case-control study. In future investigations, this sub-group should be included since vulnerable coronary lesions have a high prevalence in non-obstructive CAD [
33]. In so doing it will be possible to construct a more complete picture of the association between the structure and composition of the carotid artery and the angiographic severity of coronary disease. In addition, a prospective study including the full spectrum of CAD that aims to identify features or combinations of features in the carotid artery that predict future events should be conducted to establish the utility of this technology for risk stratification.
Conclusion
We believe our study provides evidence for several conclusions. First, high spatial resolution CMR identifies distinct differences in both morphology and composition in the carotid arteries between patients with angiographically documented obstructive CAD and disease-free controls. Second, the distal bulb and ICA of males may be more susceptible to atherosclerotic disease than the common carotid or any segment of the carotid artery of females. Finally, the carotid artery may contain multiple indicators of CAD status that may prove useful during future investigations.
Acknowledgements
This work was supported by grants from the NIH: Carotid Atherosclerosis (MRI) Progression Study (CAMPS, HL076378) and Cardiovascular Research Training Program (T-32, HL07838); and the General Clinical Research Center at the Wake Forest University School of Medicine (M01 RR-07122).
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
HRU was in charge of the MR image review, collection and interpretation of data, and preparation of the manuscript. CY is a co-investigator of this study's grant and was in charge of CMR image protocol development and validation. He also assisted in data interpretation and manuscript revision. JGT assisted in development of the study design and interpretation of data. He also significantly revised the manuscript. HC, MAE and RT performed the primary analysis of the data and assisted in data interpretation. They also contributed to the drafting of the manuscript, particularly the section on statistical analysis. TSH assisted in data interpretation and was key to revising the critical content of the manuscript. TS, BC, NT, WY, and MO reviewed and peer-reviewed MR images. In addition, they assisted in data interpretation and manuscript revision. VY and RK assisted in image protocol development and study design. They also contributed primarily to the manuscript, particularly in the section describing the CMR protocol. JJC and JM were involved with study design and revision of the manuscript. JRC is the primary investigator of this study's grant. He was involved in all parts of this study and manuscript. All authors have read and approved submission of this manuscript. The material in the manuscript has not been published and is not being considered for publication elsewhere in whole or in part in any language.