Morphological and Functional Assessment of Carotid Plaques Have Similar Predictive Accuracy for Coronary Artery Disease
Background and Purpose—Microwave radiometry allows noninvasive in vivo measuring of internal temperature of tissues reflecting inflammation. In the present study, we evaluated the predictive accuracy of this method for the diagnosis of coronary artery disease (CAD).
Methods—Consecutive patients (n=287) scheduled for coronary angiography were included in the study. In carotid arteries of both groups, the following measurements were performed: (1) intima-media thickness (IMTmax) and (2) temperature measurements by microwave radiometry (ΔTmax). C-statistic and net reclassification improvement were used to compare the prediction ability of the markers IMTmax and ΔTmax for the presence of CAD and multivessel CAD.
Results—Of 287 patients, 239 had stenoses ≥50% (CAD group), and 48 did not have significant stenoses (NO-CAD group). ΔTmax was an independent predictor for the presence of CAD and multivessel CAD, showing similar predictive accuracy to intima-media thickness, as assessed by c-statistic and net reclassification improvement.
Conclusions—Local inflammatory activation, as detected by microwave radiometry, has similar predictive accuracy to intima-media thickness for the presence and extent of CAD.
Carotid intima-media thickness (IMT) has been widely used as a surrogate marker for coronary artery disease (CAD).1–4 The increase of IMT in carotid arterial wall reflects morphological alterations and does not provide information on the functional changes. Microwave radiometry (MR) is a new method and allows in vivo noninvasive measurement of the temperature of carotid atherosclerotic plaques, reflecting their inflammatory status.5–7
In the present study, we aimed to (1) examine whether the inflammation of carotid artery wall, as assessed by MR, correlates with the presence of CAD, and (2) compare the predictive value of the morphological and functional assessment of carotid plaques in CAD detection.
Consecutive patients undergoing coronary angiography were prospectively enrolled in the study. Significant CAD was defined as diameter lumen stenosis of ≥50% in ≥1 major coronary artery. Multivessel CAD was defined as the presence of significant stenoses in ≥2 major coronary arteries.
All patients underwent carotid ultrasound examination followed by MR measurements, performed by specialists blinded to angiographic results. All participants provided informed consent, and the study was approved by our institution ethical committee.
The ultrasound protocol included scanning of both carotid arteries from their point of origin throughout their whole length using a 7.5-MHz transducer (Philips iEE33; Philips, Washington). IMT measurements were performed in 3 segments of 20 mm along each carotid artery. For this purpose, the middle segment was the region of common carotid bifurcation and was used as a marker. The regions 20 mm proximal and 20 mm distal to the bifurcation region defined the 2 other segments, namely the common carotid artery segment and the internal carotid artery segment, respectively. The highest IMT value of all segments of both carotid arteries was defined as IMTmax.
RTM-01-RES, a microwave computer-based system (Bolton, United Kingdom), was used for the performance of MR measurements. The system measures temperature from internal tissues at microwave frequencies. The basic principles of MR have been described previously.5,6
Temperature measurements were performed at each carotid artery over the previously defined segments (20 mm each). The method has been validated as described previously.6 Temperature difference (ΔT) of each carotid artery was assigned as the temperature of the segment under investigation minus the minimal temperature of each carotid. ΔTmax was assigned as the maximum ΔT value of both carotid arteries.
Data were analyzed using commercially available software (version 20, SPSS Inc; Chicago, IL). P values are 2-sided from the Student t test when continuous variables were compared and ANOVA test when there were >2 categories. Noncontinuous values were compared by χ2 test. A P value of <0.05 was considered significant. Multiple logistic regression analysis was used to determine independent predictors for CAD. We considered the addition of the following markers: (1) ΔTmax, (2) IMTmax, and (3) IMTmax plus ΔTmax on established risk factors (sex, age, dyslipidemia, arterial hypertension, diabetes mellitus, smoking, and family history), and we compared the respective prediction models with the use of Harrell c-statistic and net reclassification improvement (NRI).
We included a total of 287 patients (Table 1). Of those 287 patients, 239 (83.3%) had significant CAD (CAD group), and 48 patients did not have significant coronary artery stenoses (NO-CAD group); 103 patients (35.89%) had 1-vessel CAD, 90 patients (31.36%) had 2-vessel disease, and 46 patients (16.03%) had 3-vessel disease.
Carotid arteries of patients with NO-CAD (n=96; 16.72%) had lower ΔT compared with patients with 1-vessel CAD (n=206; 35.89%), 2-vessel disease (n=180; 31.36%), and 3-vessel disease (n=92; 16.03%; 0.48±0.39°C, 0.68±0.44°C, 0.85±0.49°C, and 0.87±0.54°C; P<0.01; Figure).
In multivariate logistic regression analysis, ΔTmax was an independent predictor for the presence of CAD when adjusted for sex, age, statin use, and established risk factors (odds ratio, 3.80; 95% confidence interval, 1.27–11.37; P=0.02). Furthermore, in the same model, ΔΤmax was an independent predictor for the presence of multivessel CAD (odds ratio, 1.98; 95% confidence interval, 1.16–3.37; P=0.01).
By receiver-operating curve analysis, we obtained a good predictive capacity of ΔTmax for CAD (area under the curve, 0.766; 95% confidence interval, 0.686–0.847; P<0.01) and for multivessel CAD (area under the curve, 0.660; 95% confidence interval, 0.597–0.722; P<0.01).
In Table 2, the results of the c-statistic and NRI values of our prediction models are summarized. The NRI with the addition of ΔTmax was higher for the prediction of multivessel CAD compared with the NRI after addition of IMTmax. For the presence of CAD, NRI was greater for IMTmax compared with ΔTmax.
This study demonstrated that: (1) patients with CAD had increased local inflammation in carotid plaques compared with subjects without significant CAD; (2) this local inflammatory activation increased proportionally to the extent of CAD as detected by angiography; and (3) inflammation as detected by MR in carotid plaques was an independent predictor for the presence and extent of CAD.
In the current study population, risk prediction models, on the basis of structural or functional markers, showed similar predictive accuracy for CAD (c-statistic IMTmax, 0.880; c-statistic ΔTmax, 0.880; Pdif=0.96) and multivessel CAD (c-statistic, 0.730 and 0.716; Pdif=0.47). Furthermore, the prediction model based on ΔTmax showed higher NRI value compared with IMTmax for multivessel CAD prediction (0.577 and 0.422, respectively). Whether these findings are reproducible in low-risk asymptomatic populations should be examined further. Possibly, the functional assessment in an earlier stage of the disease could increase the predictive value of current imaging modalities.
Local inflammatory activation of carotid arteries, as assessed by MR, correlates with the presence and extent of CAD.
We thank Dr Demosthenes Panagiotakos for his substantial contribution in the statistical analysis of the data.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.002462/-/DC1.
- Received June 11, 2013.
- Accepted June 18, 2013.
- © 2013 American Heart Association, Inc.
- Amato M,
- Montorsi P,
- Ravani A,
- Oldani E,
- Galli S,
- Ravagnani PM,
- et al
- Kotsis VT,
- Pitiriga VCh,
- Stabouli SV,
- Papamichael CM,
- Toumanidis ST,
- Rokas SG,
- et al
- Toutouzas K,
- Drakopoulou M,
- Aggeli C,
- Nikolaou C,
- Felekos I,
- Grassos H,
- et al