The early 1990s represented a turning point in the diagnosis and treatment of ICA stenosis. At this point, symptomatic patients with ICA stenosis ≥ 70% were identified as those who needed surgical treatment, and the degree of stenosis became the focal point of physicians' attention. The sonographic appearance of the plaques themselves assumed secondary importance. As a result, in most clinical settings, follow-up frequency for patients with atherosclerotic carotid artery disease is determined by the degrees of stenosis, and little attention is given to plaque features.
The first prospective study to demonstrate a link between stroke and the echogenicity of carotid plaques was conducted by Johnson et al. 
. Three years of follow-up data on 297 symptomatic patients showed that TIA or strokes had occurred in 51% of those whose plaques had been hypo or anechoic at baseline, as opposed to only 4.4% whose plaques had appeared hyperechoic. One of the most interesting studies was a multicenter initiative organized by the European Carotid Plaque Study Group. Initially published in July 1995 and republished in September of 2011 [5,6]
this study included 270 patients scheduled for carotid endarterectomy in 9 different facilities in Europe. The main findings were as follows: 1) carotid plaque echogenicity on B-mode imaging was inversely correlated with the “soft” material contained in the plaque (p
= 0.005), and plaque hyperechogenicity was directly correlated with the presence of calcifications (p
< 0.0001); 2) the most recent symptoms of CVD were reported by patients with the “softest” plaques. Mathiesen et al. 
reported that in patients with type 1 or type 2 plaques but no carotid stenosis, the relative risk (RR) for cerebrovascular events was 13 (95% CI 4.5–37.4) versus only 3.7 (95% CI 0.7–18.2) in those with type 3 or type 4 plaques. Reiter et al. 
followed 574 patients for a mean of 3.2 years to determine whether those with hypoechoic or anechoic plaques were at risk for major adverse cardiovascular events (MACEs). Carotid plaque echogenicity was assessed at baseline and every 6–9 months (mean 7.5) thereafter with dedicated software for gray-scale median (GSM) analysis. The presence of a hypoechoic plaque predicted a MACE with a hazard ratio (HR) di 1.71 (95% CI 1.09–2.66) when patients in the lowest quartile were compared with those in the highest quartile. On the whole, a total of 269 MACEs were observed in 177 (31%) of the patients studied (21 MIs, 48 coronary artery angioplasty procedures, 17 coronary artery by-pass procedures, 34 strokes, 69 angioplasty procedures involving peripheral vessels, 13 surgical procedures involving peripheral vessels, 5 amputations for critical limb ischemia, and 89 deaths).
Carotid artery plaque morphology and IMT are expressions of different biological aspects of atherosclerosis with different implications in terms of vascular disease 
. The IMT is an important predictor of cardiovascular disease, but it displays closer correlation with left ventricular hypertrophy than with atherosclerotic coronary artery disease. Compared with the IMT, carotid plaque morphology and surface area are both better predictors of stroke, myocardial infarction, and cardiovascular death [10–12]
. A systematic review of the literature on cardiovascular risk stratification of asymptomatic patients 
examined numerous studies (n
) that had assessed flow-mediated vessel dilatation (FMD, n
= 2), carotid IMT (n
= 12), carotid plaques (Plaque, n
= 6), and coronary artery calcification (CAC, n
= 9). Twenty-five studies were selected for the final analysis. The authors concluded that increases in the IMT or the presence of carotid plaques provide better estimates of the level of risk than CACs in patients with low to intermediate cardiovascular risks. In a recently published study 
, the carotid IMT plus the presence/absence of carotid plaques proved to be a better marker of CVD than either of these parameters alone. Naturally, the assessment of other markers (besides the classical indices regarding management and outcomes) in asymptomatic patients has cost-benefit implications that have to be considered.
In our study, 518 (69.4%) of the 747 patients (males 261[50.4%], females 257 [49.6%]) presented ICA stenosis, and no stenosis was found in 229 (30.6%) of the patients (males 90 [39.3%], females 139 [60.7%]). The 91 patients with ICA stenosis of >50% represented 12.2% of our cohort. This is a considerably higher proportion than that (3.6%) reported by Weerd et al. 
, but they examined a sample of the general population, whereas our patients all had CVD or were at risk for CVD. Data from the studies discussed above clearly show the importance of identifying patients with type 1 or type 2 carotid artery plaques, which are associated with an increased risk of CVD. In our study, 160 (21.4%) of the patients had ICA stenosis of 1–69% with type 1 or 2 plaques. In 83 (51.9%) of these individuals the stenosis involved the RICA, in 36 (22.5%) it was confined to the LICA, and in 41 (25.6%) it was bilateral. An interesting aspect that could have important implications for the follow-up of these patients is the progression rates of these lesions 
. We know that the thickness of a carotid plaque increases 2.4 times faster than the IMT (0.0147 mm–0.0176 mm per year), but the latter increase is below the detection threshold of sonography (approximately 0.3 mm). For this reason, measurement of plaque thickness is more sensitive than IMT measurements 
conducted a retrospective analysis of 1469 patients from the deferred endarterectomy arm of the Asymptomatic Carotid Surgery Trial. He found that the rate of progression of carotid stenosis should be considered a marker of the risk of future homolateral neurological events. But in all probability, the symptoms caused by atherosclerotic plaques are not based solely on hemodynamic mechanisms related to reduction of the carotid lumen and secondary diminution of perfusion: they are also related to thromboembolic mechanisms 
. Consequently, plaques that are not associated with significant carotid stenosis or that appear to be small on sonography can also be sources of high risk, particularly if the plaque shows signs of inflammation (i.e., macrophage infiltrates), rupture of the fibrous cap, a high lipid content, or intraplaque hemorrhage, as shown by the prospective study of Takaya et al. 
, the first group to demonstrate the association between carotid plaques and cerebrovascular events using magnetic resonance imaging (MRI). Numerous reports indicate that medical therapy—in particular, aggressive statin therapy—plays a very important role in the progression of carotid stenosis and the stability of the plaques [20–22]
In light of these studies and their clinical correlations, the follow-up of patients with ICA stenosis should be based not only on the degree of stenosis (as it is in most clinical settings) but also on the sonographic morphology of the carotid plaques. For this reason, our study () included cases of ICA stenosis of <50% and those with ICA stenosis ranging from 50% to 69%, as well as plaque type classified according to the modified G-W system. This analysis identified four groups of patients—class A, class B, class C, and class D—whose follow-ups (duration in months, initial follow-up visit, subsequent follow-up visit if the sonographic picture was stable) were planned with different temporal characteristics. In particular, for class A and class B patients the interval between the initial and subsequent follow-up visits was shorter than it was for patients in class C or D, who were considered at lower risk. In our opinion, this type of approach is more consistent with current clinical knowledge, and it should ensure more appropriate control of patients with high-risk plaques.
Follow-up schedules for four patient classes defined by sonographic plaque type (Gray-Weale) and percentage of ICA stenosis.
One of the limitations that must be considered in our study is related to differentiation of the five types of plaque envisioned by the G-W system, which involves subjective judgments, especially for distinguishing between types 2 and 3. The use of software for analyzing plaque echogenicity could substantially facilitate this task although it is poorly suited for routine clinical use. Identification of type 1 plaques, which are uniformly anechoic or hypoechoic, might also represent a limitation of this study. In the latter case, proper settings (B-mode signal, gray-scale level, color signal, PRF) should be diagnostically useful. A final limitation is related to the rigidly defined temporal characteristics of the follow-up. It could be overcome by adapting the follow-up to the overall cardiovascular risk status of the patient being examined.