Carotid Artery Plaque Classification: Does Contrast Enhancement Play a Significant Role?

Study Design and Patient Population

At our institute, a carotid artery data base was created from retrospective case material, from 2004 onward, that includes all CTA examinations of carotid arteries performed at the institute.

Based on a power calculation (type I error, α = .01; type II error, β = .01; difference = .05), we estimated that a sample size of at least 400 carotid arteries would be sufficient to evaluate the effect of enhancement in carotid artery plaque classification. Therefore, we searched the data base and found 380 consecutive patients (235 men; 145 women; median age 64, age range 32–87 years) who were examined with MDCTA from May 2006 to February 2011. Part of this population was used in previous studies.¹⁰ In our institute, CTA was performed when there was a sonography (US) examination that showed pathologic stenosis (50% of stenosis, according to NASCET criteria) and/or plaque alteration (presence of heterogeneous plaque, an irregular surface, intraplaque hemorrhage, and/or the presence of ulceration in the plaque). MDCTA was also performed when the US could not provide adequate information about the degree of stenosis and plaque type. Elevated renal function tests or known allergy to iodinated contrast material were considered as exclusion criteria for the MDCTA examination. In our hospital, carotid arteries of asymptomatic patients who underwent cardiac interventions for coronary artery disease, aortic interventions, and lower leg artery surgery, and diabetics who were >50 years old, were also studied.

Institutional review board approval was obtained and this work was conducted in accordance with the guidelines of our institution's research committee. Because of the retrospective nature of this analysis, the institutional review board waived the need for ethics committee review.

MDCTA Technique

All patients underwent MDCTA of the supra-aortic vessels using 2 different CT scanners—a 40 multidetector row CT system (Somatom Sensation; Siemens, Erlangen, Germany) and a 16 multidetector row CT system (Brilliance; Philips, Best, the Netherlands)—using a technique previously described.

In our protocol for the analysis of carotid arteries, a noncontrast scan was performed and was followed by the angiographic phase. The angiographic phase was obtained by injecting 80–90 mL of contrast medium (Ultravist 370; Schering, Berlin, Germany) into a cubital vein, using a power injector at a flow rate of 5 mL/s and an 18-gauge intravenous catheter. A bolus tracking technique was used to calculate the correct timing of the scan. Scans were performed in the caudocranial direction. CT technical parameters included section thickness, 1 mm; interval, 0.5 mm; matrix, 512 × 512; FOV, 14–19 cm; mAs, 180–200; and kV, 120–140. Angiographic acquisition included the carotid siphon. None of the patients included in the study had a medical history of cardiac output failure.

Carotid Image Quality and Artifact Evaluation

Two radiologists (7 and 11 years of experience in MDCTA of the supra-aortic vessel) were asked to evaluate, in consensus, the overall image quality on a 5-point scale, where 5 corresponded to an excellent image quality, 4 to a good image quality, 3 to an adequate image quality, 2 to a marginally acceptable image quality, and 1 to an unacceptable image quality. Patients for whom at least 1 reader rated image quality as 1 or 2 were excluded from this study.

First Plaque Type Analysis (Contrast-Enhanced Phase)

Two radiologists (7 and 11 years of experience in MDCTA of the supra-aortic vessel) performed all HU measurements, blinded to each other. Window/level settings were usually set at W850:L300, progressing to very wide settings in the case of attenuated calcifications.¹¹

In the first plaque type analysis, we assessed only the contrast material dataset. A circular or elliptical region of interest (≥2 mm²) in the predominant area of the plaque was used to measure the HU value (Fig 1). Areas showing contamination by contrast material or calcification that did not contribute to the stenosis were avoided. Regions of beam hardening in calcified areas were also excluded. Moreover, a note was made of the corresponding section to identify any differences between the observers in the choice of the plaque along the z-axis.

• Download figure
• Open in new tab
• Download powerpoint

Fig 1.

Plaque analysis. In panel A, the basal phase is given, whereas panel B is the phase after administration of contrast material. The white arrows indicate the carotid artery plaque, whereas the red circle is the region of interest.

Second Plaque Type Analysis (Precontrast Phase)

The observers first selected the precontrast baseline MDCTA dataset and identified the section corresponding to the contrast enhancement (Fig 1). To obtain a correct registration between the 2 datasets, they visually assessed whether the sections corresponded or whether it was necessary to select another section along the z-axis. After this “matching phase,” to measure the basal HU value, a region of interest of the same area as that used in the contrast phase was placed in the same position on the plaque.

Plaque-Type Classification and Interobserver Analysis

As in previous studies,^6,12 plaques were classified as 1) fatty—plaques with attenuation <60 HU; 2) mixed—plaques with attenuation between 60 and 129 HU; and 3) calcified—plaques with attenuation >130 HU. This classification was based on the precontrast scan.

Statistical Analysis

To test the differences between plaque classification before and after administration of contrast material, a χ² 3 × 2 test for differences was used. To perform the analysis, the mean values between the 2 observers were considered. A Student t test for independent samples was used to compare the plaque enhancement in fatty plaques that did not change their category to mixed type versus the plaque enhancement in those fatty plaques that changed their type. Interobserver reliability for plaque classification was calculated using the Cohen κ test. The following interpretation of the κ statistic was used: 0–0.20 = poor agreement; 0.21–0.40 = fair agreement; 0.41–0.60 = moderate agreement; 0.61–0.80 = good agreement; and 0.81–1 = very good agreement. To produce these data, R software (www.r-project.org) was used. A P value <.05 was regarded as indicating statistical significance.