Background
Although carotid intima–media thickness (CIMT) assesses the structural properties of the carotid artery, it does not assess the mechanical properties of the vessel.
Methods
The carotid arteries of 71 adult patients were evaluated with CIMT, and automated border detection computed vessel stiffness, compliance, elasticity, and distensibility.
Results
CIMT and mechanical properties were differentially affected by traditional cardiac risk factors, with age dominating for CIMT, and age, diabetes, and smoking dominating for mechanical variables. There was a moderate linear correlation between CIMT and the distensibility coefficient ( r = −0.54), but there were weak associations with other parameters of dynamic vessel function. When patients were separated into risk groups, the mechanical vascular parameters’ classification frequently differed from the CIMT classification. This was particularly notable for patients with intermediate CIMT values, who were reclassified as low or high risk by mechanical parameters 45% of the time.
Conclusion
We found that it is feasible to assess the cross-sectional area of the carotid artery using automatic border detection, which allows a novel method of determining carotid mechanical properties. These functional characteristics are often discordant with CIMT, suggesting that mechanical properties may be an important adjunct to the CIMT when evaluating the carotid artery.
Noninvasive assessment of atherosclerosis is important to gauge the risk of stroke, coronary artery disease, and peripheral vascular disease. The common carotid arteries are easily accessible by ultrasound, and carotid intima–media thickness (CIMT) has emerged as a useful predictor of cardiovascular risk; however, although CIMT assesses the structural properties of the vessel, it does not provide information about the mechanical properties of the vessel. Although increases in CIMT and deterioration in carotid mechanical properties have similar risk factors, studies evaluating the relationship between structural and mechanical properties of the vessels have shown variable results.
The mechanical properties of the carotid arteries have been evaluated by measuring parameters such as vessel stiffness, compliance, elasticity, and distensibility. Assessment of these mechanical properties is typically done by measuring linear arterial diameter in combination with arterial pressure. Acquisition of carotid vessel dimension requires either manual measurement or specialized wall tracking software. However, the pulsatile distension of a blood vessel is best assessed by measuring change in vessel cross-sectional area rather than measuring linear dimension. Accurate tracking of changes in vessel dimension requires that the ultrasound beam remain on-axis through the center of the vessel while the vessel pulsates. In addition, the change in cross-sectional area represents the properties of the entire vessel at that level rather than just a single point on the vessel wall, and is therefore less affected by circumferential regional variations in vessel wall properties. Because continuous area measurement would require tedious frame-by-frame tracing throughout systole and diastole, mechanical assessment of the carotid using changes in area has not been performed.
Automated border detection (ABD) with acoustic quantification (AQ) is an echocardiographic technique that identifies the endocardial-blood pool border and displays continuous chamber size throughout the cardiac cycle. This technique has been used to accurately assess left ventricular and left atrial chamber cross-sectional area. AQ has also been used to track the endothelial-blood pool border in larger vascular structures, such as the descending thoracic aorta. We sought to investigate 1) whether AQ can be used to determine pulsatile carotid cross-sectional area and calculate parameters of arterial mechanical function; 2) how AQ-determined mechanical properties of the carotid artery compare with traditional structural assessment of the carotids with CIMT; and 3) how AQ-determined mechanical properties of the carotid artery and CIMT are influenced by vascular risk factors.
Materials and Methods
Seventy-one consecutive adult patients referred for transthoracic echocardiographic examinations at the University of Chicago were recruited. Exclusion criteria included atrial fibrillation or history of a carotid endarterectomy on the left side. Informed consent was obtained from each patient in accordance with the institutional review board at University of Chicago. Patient demographics and vascular risk factors—hypertension, hypercholesterolemia, diabetes, smoking, and age (50 years for men, 60 years for women)—were determined by interview and chart review. Systolic (Ps) and diastolic (Pd) blood pressures were measured from the left brachial artery with the patient in the supine position immediately before acquisition of carotid images. Patient medications were not held or altered during the study.
Images were acquired by sonographers trained in carotid imaging using an iE33 ultrasound system (Phillips Medical Systems, Andover, MA). For acquisition of the cross-sectional view of the common carotid, a 2-dimensional short-axis view was taken 1 to 2 cm below the bulb of the left carotid artery using a linear (L11-3) probe. Care was taken to maintain an orthogonal true short axis, and 10 beats were acquired and digitally stored. To determine carotid cross-sectional area throughout the cardiac cycle, semiautomated software (QLab, Phillips Medical Systems) was used to identify the blood pool–endothelial interface ( Figure 1 ). Adequate border identification was visually assessed in real-time, and manual adjustments in gain could be used to improve endovascular border tracking. An averaged curve of cross-sectional area over the 10 beats was generated ( Figure 2 ). From the averaged area curve, end-systolic area (As) and end-diastolic area (Ad) were computed by the software. End-systolic dimension (Ds) and end-diastolic dimension (Dd) were derived assuming a circular vessel. As a validation of the ABD tracking, the end-diastolic and end-systolic areas of the carotid were hand-traced on two-dimensional images with border tracking suppressed, but all other images settings unchanged, in 10 patients.
The carotid was scanned for plaques using both transverse and longitudinal planes. Patients with visible atherosclerotic plaque were excluded from the study. Images for CIMT assessment were acquired and optimized independently from those used for ABD analysis. To acquire images for measurement of CIMT, a longitudinal view was taken of the lateral and posterior walls of the distal 1 cm of the common carotid for a continuous loop of 6 beats using high-resolution B-mode ultrasound. Breath-holding was used only if there was motion artifact. An experienced reader measured the mean CIMT of the common carotid artery using ABD software (QLab, Phillips Medical Systems) over a 10-mm sample at end diastole.
Parameters used to measure the mechanical properties of the carotid artery included beta-stiffness (β) = ln(Ps/Pd)/[(Dd-Ds)/(Dd)]; cross-sectional compliance (CSC) = (Dd-Ds)/(Ps-Pd); distensibility coefficient (DC) = (Ad-As)/[Ad(Ps-Pd)]; Peterson’s elastic modulus (EM) = [(Ps-Pd)Dd]/(Dd-Ds); and Young’s modulus = [(Ps-Pd)Dd 2 ]/[2(Dd-Ds)CIMT].
Statistics
Associations between the structural and mechanical properties of the carotid were assessed with correlation analysis. Patients were categorized into quartiles using each variable. The second and third quartiles were combined to make 3 groups for each parameter: low-risk (1st quartile) and high-risk (4th quartile) groups with a larger intermediate-risk group (middle 2 quartiles). The risk group classifications were compared between the structural and mechanical parameters. All the structural and mechanical variables were evaluated with univariate analysis of variance using the 5 major risk factors as categoric inputs. Statistical significance was considered as a P value less than .05.
Results
Of the 71 patients, 29 were male and 42 were female with an average age of 57 ± 20 years. Forty percent of women were premenopausal. Baseline demographic data are shown in Table 1 . There was an average of 2.2 (of 5) vascular risk factors per patient, and 31% had established vascular disease (cardiac or peripheral). ABD carotid area and CIMT measurements were possible in all patients; the results are shown in Table 2 . There was excellent agreement between hand-traced and ABD carotid cross-sectional area (0.44 ± 0.11 cm 2 , 0.45 ± 0.10 cm 2 , bias 0.01 cm 2 , standard deviation of differences 0.03 cm 2 ).
| Age (y) | 57 ± 20 |
| SBP (mm Hg) | 129 ± 20 |
| DBP (mm Hg) | 71 ± 12 |
| Hypertension | 65% |
| Hyperlipidemia | 34% |
| Diabetes | 27% |
| Smoker | 30% |
| Known vascular disease | 31% |
| Mean ± SD | Correlation with age | Correlation with CIMT | Univariate risk factor predictors | |
|---|---|---|---|---|
| CIMT (mm) | 0.72 ± 0.20 | 0.61 | – | Age |
| Ad (cm 2 ) | 0.40 ± 0.17 | – | ||
| As (cm 2 ) | 0.44 ± 0.18 | – | ||
| β | 12.9 ± 9.6 | 0.41 | 0.38 | DM |
| CSC (mm 2 /kPa) | 0.68 ± 0.41 | −0.53 | −0.39 | Smoking, DM, age |
| DC (10 −3 /kPa) | 19.0 ± 11.9 | −0.67 | −0.54 | Age |
| Peterson’s EM (kPa×10 3 ) | 0.17 ± 0.13 | 0.40 | 0.36 | DM |
| Young’s modulus (kPa×10 3 ) | 0.85 ± 0.80 | 0.26 | 0.12 | – |
When evaluated against age, there were both structural (increase in CIMT) and mechanical alterations in the carotid artery ( Table 2 ). The strongest correlations with age were for CIMT and the distensibility coefficient. When considering the traditional risk factors for atherosclerosis, although there were trends for worsening vascular function with more risk factors, a significant difference was only noted for CIMT and distensibility between groups with 0 to 1 and 2 to 3 risk factors ( Table 3 ). When comparing the lowest (0-1 risk factors) and highest (>3 risk factors) risk groups, most parameters differed between these patients ( Table 3 ). Different parameters were affected variably by the 5 major risk factors. Age was a statistically significant predictor for CIMT, CSC, and DC. The presence of diabetes affected the beta-stiffness (β), CSC, and Peterson’s EM. CSC was also influenced by smoking status.
| Risk factors | 0-1 (n=19) | 2-3 (n=31) | >3 (n=21) |
|---|---|---|---|
| CIMT (mm) | 0.54 ± 0.10 | 0.76 ± 0.20 a | 0.81 ± 0.18 a |
| β | 9.0 ± 7.7 | 12.6 ± 6 | 17 ± 13 a |
| CSC (mm 2 /kPa) | 0.84 ± 0.42 | 0.68 ± 0.43 | 0.53 ± 0.31 a |
| DC (10 −3 /kPa) | 27 ± 12 | 17 ± 10 a | 16 ± 11 a |
| Peterson’s EM (kPa×10 3 ) | 0.11 ± 0.11 | 0.17 ± 0.08 | 0.22 ± 0.19 a |
| Young’s modulus (kPa×10 3 ) | 0.69 ± 0.67 | 0.80 ± 0.42 | 1.1 ± 1.2 |
a Significantly different than the group with 0 to 1 risk factors.
There was a moderate correlation between CIMT and the mechanical properties of the carotid as assessed by the distensibility coefficient ( r = −0.54), but fairly weak associations between carotid wall thickness and other parameters of dynamic vessel function ( Table 2 , Figure 3 ). When patients were separated into risk groups, the classifications of the mechanical vascular parameters and CIMT frequently differed. Table 4 shows the risk group distributions by CIMT and arterial properties. Patients with the highest values of CIMT had a top risk group classification from the mechanical variables only 22% to 56% of the time. For patients with the lowest CIMT values, their classification was identical 39% to 56% of the time with vascular parameters. Finally, those patients in the intermediate CIMT group were frequently categorized in the highest (20%-34%) or lowest (17%-26%) risk group by the mechanical parameters.
