Cardiovascular risk factors and cervical carotid plaque features on CT angiography

Abstract

Background

Little is known about the impact that cardiovascular (CV) risk factors have on the formation of various carotid atherosclerotic plaque features. This study set out to assess the association between CV risk factors and plaque characteristics on computed tomography (CT) angiography (CTA).

Materials and methods

A retrospective review was completed of consecutive patients that underwent a carotid endarterectomy and had CTA imaging of the head and neck vasculature. Atherosclerotic plaques of both carotid arteries were evaluated for calcification(s), low-density plaque (LDP) components, ulceration(s), and degree of stenosis. Various clinical CV risk factors were assessed using medical records. Last recorded laboratory levels were dichotomized into categories: total cholesterol <200 or ≥200 mg/dL, low-density lipoprotein (LDL) <130 or ≥130 mg/dL, high-density lipoprotein <35 or ≥35 mg/dL, and triglyceride <200 or ≥200 mg/dL.

Results

Of 97 included patients, 62 were male (63.9%); the average age was 72.7 (standard deviation = 9.5). Calcifications were in 95/97 (97.9%) of patients (one or both carotid plaques); LDP components were in 73/97 (75.3%), and ulcerations were in 21/97 (21.6%). Elevated total cholesterol and elevated LDL levels were both associated with a higher likelihood of LDP components (p = 0.004 and p = 0.02, respectively). There were no other statistically significant associations between individual plaque features or severity of arterial stenosis and CV risk factors.

Conclusion

In carotid atherosclerotic plaques, LDP components are more frequently present in one or both carotid arteries in patients with elevated total cholesterol and/or LDL levels. Such findings raise the possibility that cholesterol levels may be directly related to the formation of specific high-risk plaque features.

Introduction

Over the past decade, there has been substantially increased interest in the imaging features of carotid artery atherosclerosis. The majority of such studies have focused on the use of magnetic resonance imaging (MRI), while computed tomography angiography (CTA) has remained mainly relegated to assessing the degree of arterial stenosis.^1,2 Nevertheless, CTA is a well-established tool for the assessing carotid atherosclerosis. ³ Like MRI, CTA can be used to characterize multiple plaque features.^4,5 And, because the use of CTA is nearly ubiquitous in stroke and vascular imaging, it remains an enticing modality for characterizing atherosclerotic plaques.

Plaque features on CTA have both diagnostic and prognostic value: calcified plaques are more stable, while plaques with low-density plaque (LDP) components or ulcerations are more vulnerable to acute changes and ipsilateral ischemic events.^6,7 Understanding the risk factors for developing high-risk features may be useful to identify patients at increased risk of stroke. To date, however, there is little known about the cardiovascular (CV) risk factors that contribute to the development of various carotid plaque features on CTA. This study sought to assess whether any such associations exist between plaque features on CTA and CV risk factors in a cohort of patients that underwent carotid endarterectomy (CEA).

Materials and methods

Patient selection

Following approval from the institutional review board, a retrospective review was completed of consecutive patients that underwent CTA imaging of the cervical arterial vasculature between 10/25/2002 and 2/7/2020. Included patients underwent carotid CEA following the CTA, with atherosclerotic plaque tissue available for histologic analysis. Concordance between imaging and histology is anticipated to be published in a future manuscript, and will not be reviewed here. Patients were excluded if: (a) surgery was not performed or (b) images were substantially degraded by artifact (e.g. motion). If multiple CTAs were available for analysis, the imaging performed in closest temporal relation to the CEA surgery – but not after the surgery – was used.

CTA imaging protocol

The majority of CTA imaging was completed on a 128-slice multi-detector scanner (SOMATOM Definition Flash; Siemens Healthcare, Erlangen, Germany). Scanning was performed from the carina to the vertex. Omnipaque 350 (GE Healthcare, Mississauga, Canada) was administered at a rate of 4 mL/s (total volume of 100 mL) followed by a flush of normal saline at 4 mL/s (total volume 35 mL). Contrast administration was automatically triggered by bolus tracking through the aortic arch. Slice thickness was 0.75 mm, FOV 250 mm, pitch 0.6. Tube voltage (kV) and tube current (mAs) were set to 120 kV and 294 mAs, respectively.

Imaging analysis

All CTA examinations were reviewed by a single neuroradiologist (JCB) who was blinded to the CV risk factors and subsequent surgery. Atherosclerotic plaques of both the right and left carotid arteries were evaluated for the presence or absence of (a) calcifications, (b) LDP components, and (c) ulcerations. LDP components were defined as having mean internal Hounsfield units of less than 60 HU. ⁷ Ulcerations were defined as being excavations into the plaque, allowing contrast to extend beyond the vascular lumen for at least 1 mm. ⁸ Calcifications were defined as being >130 HU. The degree of stenosis was calculated using North American Symptomatic Carotid Endarterectomy Trial (NASCET) criteria. ⁹ The degree of calculated arterial stenosis was categorized into mild (<50%), moderate (50–69%), and severe (≥70%). If discrepancies between arterial stenosis categories between sides was noted, the artery with the greater severity of stenosis was used for statistical analysis.

Electronic medical record (EMR) review

Demographic information and CV risk factors were evaluated via a retrospective review of the patients’ EMRs. Data extracted from each patient (when available) included body mass index (BMI), a history of alcohol and/or tobacco use (current or prior), known clinical history of diabetes mellitus (DM) or hypertension (HTN), last recorded cholesterol levels (total, low-density lipoprotein (LDL), and high-density lipoprotein (HDL)), and last recorded triglyceride (TG) level. The majority of laboratory values were collected at the time of CEA. Cholesterol and TG levels were dichotomized into categories: total cholesterol either <200 or ≥200, LDL <130 or ≥130, HDL <35 or ≥35, and TG <200 or ≥200 (all reported as mg/dL). All patients were also assessed for any history of prior stroke.

Statistical analysis

Means and standard deviations of all continuous variables were calculated. Statistical analyses were performed with the statistical analysis software (SAS)-based statistical software package JMP 13.0 (www.jmp.com, Cary, NC, USA). Each demographic data point and CV risk factor were compared to whether each carotid plaque characteristic (i.e. presence or absence of ulceration, calcification, or LDP) was present in either artery. Data from both arteries were used to assess whether the categorized degree of stenosis (mild, moderate, severe) was associated with patient demographics or CV risk factors. Student’s t-test was used to calculate continuous variables and chi-squared tests were used to examine categorical variables. The threshold for statistical significance was set to p < 0.05.

Results

Patient characteristics

A total of 97 patients were included in this study, of which 62 were male (63.9%); the average age was 72.7 ± 9.5 years. Of the 97 patients, 76 (78.4%) had a stroke prior to undergoing CEA, with mean time between CEA and CTA 252.0 days; 34/97 (35.1%) patients had never used tobacco, while 55 were former users and eight were current users. Therefore, 63/97 (64.9%) were either current or former tobacco users. Forty-five (46.4%) patients used alcohol. The average BMI of all patients was 29.9 ± 5.8. Laboratory values were not available for all patients. Of those with recorded data, 15/85 (82.4%) had total cholesterol ≥200 mg/dL, 10/86 (11.6%) had LDL ≥130 mg/dL, 13/85 (15.3%) had HDL <35 mg/dL, and 12/85 (14.1%) had TG ≥200 mg/dL.

Plaque characteristics

Calcifications were present in 88/97 (90.7%) of right carotid arteries, and in 91/97 (93.8%) of left carotid arteries. In each given patient, a calcification was present in either (or both) carotid artery in 95/97 (97.9%) of patients. LDP component(s) were present in 50/97 (51.5%) of right carotid arteries, and in 49/97 (50.5%) of left carotid arteries. LDP was present in either (or both) carotid artery in 73/97 (75.3%) of patients. Ulcerations were the least commonly identified, and were present in 12/97 (12.4%) of right carotid arteries and 11/97 (11.3%) of left carotid arteries, overall identified in either (or both) artery in 21/97 (21.6%) of patients (Figure 1).

Figure 1.

Example of a plaque with various imaging features in a 57-year-old male. Axial (a) and reformatted oblique coronal (b) CTA images of the neck demonstrate an atherosclerotic plaque in the proximal right carotid artery. The plaque has dominant LDP components (straight arrows), with mean density of 44.5 HU. Also noted is a small, thin ulceration (curved arrows) and a tiny focus of calcification (dashed arrow). The patient’s total cholesterol and LDL were both elevated.

Among right carotid arteries, 43 (44.3%) had mild narrowing, 21 (21.6%) had moderate narrowing, and 33 (34.0%) had severe stenosis. Among left carotid arteries, 44 (45.4%) had mild, 25 (25.8%) had moderate, and 28 (28.9%) had severe narrowing. For statistical purposes, if a discrepancy between sides was present, the more severe stenosis was used. Hence, for all patients, the greatest degree of narrowing was mild in 10/97 (10.3%), moderate in 30/97 (30.9%), and severe in 57 (58.8%) of patients.

Association of plaques with CV risk factors

The findings are outlined in Table 1. A statistically significant association was noted between the presence of LDP components and cholesterol levels: both total cholesterol ≥200 mg/dL (p = 0.004) and LDL ≥130 mg/dL (p = 0.02) were associated with a higher likelihood of LDP components. In addition, an association was noted between male gender and the presence of calcifications within a carotid plaque (p = 0.04), though this may have been influenced by the low number of plaques without calcifications. There were otherwise no statistically significant associations between the presence of carotid artery plaque LDP components, calcification, or ulcerations and CV risk factors. In addition, there was no association between CV risk factors and the degree of atherosclerotic carotid stenosis.

Table 1.

Correlations between imaging plaque features and CV risk factors. Degree of stenosis refers to the more severe stenosis in a given patient if a discrepancy between right and left arteries was present. Variations in reported numbers are related to some data not being available for each patient during EMR review. Statistical significance was noted in terms of association between total cholesterol and LDL and the presence of LDP, as well as the presence of calcification and male gender (bolded). There was a trend toward significance in terms of greater degree of stenosis and DM.

	LDP			Calcification			Ulceration			Stenosis
	+	−	P-value	+	−	P-value	+	−	P-value	<50%	50–69%	≥70%	P-value
Gender (% male)	46/73 (63.0%)	16/24 (66.7%)	0.75	62/95 (65.2%)	0/2 (0.0%)	0.04	12/21 (57.1%)	50/76 (65.8%)	0.47	4/10 (40.0%)	20/30 (66.7%)	38/57 (66.7%)	0.27
DM	19/73 (26.0%)	7/24 (29.2%)	0.76	26/95 (27.4%)	0/2 (0.0%)	0.26	5/21 (23.8%)	21/76 (27.6%)	0.72	1/10 (10.0%)	5/30 (16.7%)	20/57 (35.1%)	0.07
HTN	57/73 (78.1%)	19/24 (79.2%)	0.91	75/95 (78.9%)	1/2 (50.0%)	0.37	18/21 (85.7%)	58/76 (76.3%)	0.34	6/10 (60.0%)	25/30 (83.3%)	45/57 (78.9%)	0.33
Tobacco	48/73 (65.8%)	15/24 (62.5%)	0.77	62/95 (65.3%)	1/2 (50.0%)	0.66	12/21 (57.1%)	51/76 (67.1%)	0.40	6/10 (60.0%)	23/30 (76.7%)	34/57 (59.6%)	0.26
EtOH = Alcohol	33/73 (45.2%)	12/24 (50.0%)	0.68	44/95 (46.3%)	1/2 (50.0%)	0.92	8/21 (38.1%)	37/65 (48.7%)	0.39	2/10 (20.0%)	15/30 (50.0%)	28/57 (49.1%)	0.19
Total cholesterol ≥200 mg/dL	15/66 (22.7%)	0/19 (0.0%)	0.004	15/84 (17.9%)	0/1 (0.0%)	0.53	4/19 (21.5%)	11/66 (16.7%)	0.66	1/10 (10.0%)	7/25 (28.0%)	7/50 (14.0%)	0.28
LDL ≥130 mg/dL	10/67 (14.9%)	0/19 (0.0%)	0.02	9/84 (10.7%)	1/2 (50.0%)	0.17	2/19 (10.5%)	8/67 (11.9%)	0.86	0/10 (0.0%)	3/25 (12.0%)	7/51 (13.7%)	0.26
HDL <35 mg/dL	11/66 (16.7%)	2/19 (10.5%)	0.50	13/84 (15.5%)	0/1 (0.0%)	0.56	3/19 (15.8%)	10/66 (15.2%)	0.95	1/10 (10%)	4/25 (16.0%)	8/50 (16.0%)	0.87
TG ≥200 mg/dL	9/66 (13.6%)	3/19 (15.8%)	0.81	12/84 (14.3%)	0/1 (0.0%)	0.58	4/19 (21.1%)	8/66 (12.1%)	0.34	1/10 (10.0%)	5/25 (20.0%)	6/50 (12.0%)	0.60
BMI (mean)	29.9	29.9	0.98	30.0	22.9	0.09	29.1	30.1	0.51	29.2	30.4	29.7	0.84

Assessment of plaque characteristics versus a history of prior stroke is shown in Table 2. Overall, no such association was observed. There was a slight trend toward a higher proportion of LDP components in plaque among patients with a prior stroke (78.4%) versus those without a cerebrovascular accident (CVA) history (61.9%), though this also did not meet statistical criteria for significance (p = 0.12).

Table 2.

Comparison of various plaque imaging features to prior cerebrovascular accident (CVA). No statistically significance was observed. However, there was a trend toward significance in terms of the presence of LDP in a plaque and likelihood of a prior CVA.

	Prior stroke
	+	−	P-value
Calcification	74/76 (97.4%)	21/21 (100.0%)	0.32
LDP	60/76 (78.4%)	13/21 (61.9%)	0.12
Ulceration	17/76 (22.4%)	4/21 (19.0%)	0.74
Stenosis
<50%	9/76 (11.8%)	1/21 (4.8%)	0.35
50–69%	25/76 (32.9%)	5/21 (23.8%)
≥70%	42/76 (55.3%)	15/21 (71.4%)

Discussion

This study set out to assess whether associations exist between CV risk factors and various plaque imaging features on CTA in a cohort of patients that underwent CEA. The results indicate that elevated total cholesterol and LDL are associated with the presence of LDP plaque components. Cholesterol levels could potentially be related to the development of “soft” plaques, which are at increased risk for ipsilateral ischemic events. Other CV risk factors, conversely, may have little effect on the pathogenesis of specific plaque imaging features.

One alluring prospect raised by these results is that CTA could be used as an imaging modality to assess for, and follow, LDP components in patients with hyperlipidemia. Already, it is well-documented that the size of lipid-rich necrotic cores (LRNCs) within plaques can stabilize or regress during treatment with statins. ¹⁰ It is thought that the lipid material is replaced with other histologic elements such as fibrous tissue. ¹¹ As such, it is not surprising that an association with higher cholesterol level is observed in patients with LDP components, which are thought to partially represent lipid material. However, MRI components are often made up of a combination of LRNC and intraplaque hemorrhage (IPH). Although multiple prior efforts have been made to differentiate between these components, they have shown inconsistent results.^12,13 As such, most studies regarding delipidation of plaques during statin therapy have used MRI (or, infrequently, ultrasound) to assess plaque features.^14,15

Prior studies on the topic of CV risk factors and imaging features of plaques have had variable results. Using magnetic resonance angiography (MRA), Wasserman et al. found that lipid cores were more likely to be present in patients with elevated total cholesterol. ¹⁶ Those results are similar to what is reported in the current study, although CTA is unable to distinguish between IPH and LRNC, as mentioned. Using CTA, Li et al. assessed the relationship between carotid artery plaque features and CV risk factors. ¹⁷ In this study, the authors compared plaque imaging characteristics to the atherosclerotic CV disease (ASVD) risk, which is based on gender, age, race, cholesterol levels, smoking status, and DM. The authors found that patients with an elevated ASCVD score were more likely to have arterial stenosis, had larger maximal plaque thickness, and were more likely to have “soft” plaque components. However, the authors did note that the concordance between plaque features and ASCVD scores was imperfect. The patient cohort used by Li et al. was substantially smaller than the current study, and excluded patients that had undergone a CEA. Nevertheless, the agreement between the current and prior studies regarding LDP/soft plaque components suggests that the observed association is not limited to CEA patients.

Rozie et al. also examined the associations between CV risk factors and plaque characteristics on CTA. ¹⁸ The authors categorized presumed plaque histology based on intralesional density: calcification >130, fibrous tissue 60–130, and lipid core <60. The results of that study indicated that both age and tobacco use were associated with greater plaque volume. However, contrary to the current study, the authors found that high cholesterol levels were associated with less lipid material. It is possible that the observed differences may be related to variations in the cohorts used or in the categorization of plaque features. Nevertheless, future studies will likely be needed for further investigation.

In addition, Rovella et al. performed a study on symptomatic and asymptomatic patients with at least 60% stenosis that underwent CEA. ¹⁹ The authors dichotomized patients into having either “stable” or “unstable” plaques, based on the presence or absence of either thrombosis/thrombi or thing-cap fibroatheromas on histology. Obesity was found to be an independent risk factor for histologically unstable plaques, particularly in patients under 70 years old. However, the study did not assess specific CTA imaging features. Although the current study did not find any such association between BMI and plaque features, the vulnerable characteristics assessed for in this study (i.e. LDP components and ulceration) do not neatly conform to the histologic assessments performed by Rovella et al. As such, it is possible that obesity is a risk factor for certain histologic plaque characteristics but not plaque imaging features.

The current study has limitations shared by all retrospective analyses. In addition, the cohort was made entirely of patients that underwent CEA. It is, therefore, possible that a selection bias influenced both the observed imaging characteristics of plaques and the incidence of CV risk factors. Also, because all patients underwent a CEA, the expected degree of stenosis among the studied carotid arteries is presumably higher than a normal population. Were other patients to have been included, it is possible that a statistically significant association would have been noted between stenosis severity and one or more CV risk factors. Next, only a single neuroradiology reviewer was used to analyze the images in this study, thereby potentially introducing reader bias to the results. In addition, because no timeframe specifications were placed on the laboratory values evaluated, it is possible that variability existed in the degree Finally, the high prevalence of calcified plaque components in the cohort also limited the usefulness of statistical analysis regarding any associations, decreasing the certainty between the observed association between male gender and calcifications.

Conclusions

LDP plaque components are more commonly seen in one or both carotid arteries in patients with elevated total cholesterol and/or LDL levels. CV risk factors otherwise have no observed associations with various plaque imaging features on CTA.