What does intima???media thickness tell us?

Increased intima–media thickness (IMT) from ultrasound measurements of the carotid artery is considered as a surrogate marker of more generalized atherosclerosis [1], and has repeatedly been shown to predict the occurrence of cardiovascular events such as stroke and myocardial infarction [2–6]. Increased carotid IMT helps to better classify the cardiovascular risk in hypertensive patients [7]. Therefore, carotid IMT is being used increasingly as a marker of target organ damage in patients with increased cardiovascular risk such as hypertensive subjects [7]. The 2003 European Society of Hypertension–European Society of Cardiology guidelines recommend IMT measurements particularly in those patients in whom target organ damage is not discovered by routine investigations, including an electrocardiogram [8]. The pathophysiological concept behind carotid IMT as a marker of target organ damage is that intimal thickening at the carotid artery may be an early stage of atherosclerotic disease [9]. In vivo, intimal thickening per se cannot be measured non-invasively, but the intima–media complex can be measured by ultrasound. An increased IMT can be due to intimal and/or muscular thickening. By contrast to the more peripheral muscular arteries, the carotid artery is an elastic artery and the muscular media is relatively small. Therefore, increased carotid IMT is considered to represent mainly intimal thickening [1]. This is presumed to explain the observed association between increased carotid IMT and cardiovascular events. Is every intima–media thickening a marker of atherosclerosis? The process of intima–media thickening is complex. Intima–media thickening can be reactive to a higher blood pressure and changes in shear stress pattern. The latter explains the prominent IMT changes at the carotid bifurcation where turbulence is occurring. Reactive changes are not per se a marker of early atherosclerosis. The Rotterdam study supports the view that, at lower degrees of IMT, thickening appears to reflect an equilibrium state in which the effects of pressure and flow on the arteries are in balance, given the characteristic relationship between shear stress and local transmural pressure. Beyond a certain level, IMT more likely may represent atherosclerosis [10]. Large agreement exists that this level is around 900 μ [8]. Increased IMT and its progression is associated with well-known cardiovascular risk factors such as age, gender, body mass index, smoking, blood pressure and total cholesterol [11]. Different ultrasound techniques have been used to measure IMT of the carotid artery. In general, the different methods have acceptable accuracy and reproducibility for measuring carotid IMT, especially when applied in large multicentre clinical trials [12]. A large majority of studies on IMT have used measurements in B-mode, but IMT is also measured in M-mode or assessed from the radiofrequency signals of echotracking devices. The B-mode technique calculates the average IMT of an arterial segment [10,12,13]. Computer-assisted tools have been developed to facilitate these calculations. By contrast to the B-mode technique, the M-mode technique measures the IMT at a discrete arterial position [14] and can easily assess IMT at a fixed timepoint in the cardiac cycle: diastolic IMT. Acceptable agreement exists between the B-mode and M-mode methods without important systematic difference [12]. In recent years, echotracking devices have also been used to measure IMT. These devices can calculate IMT at a fixed timepoint and also the average IMT during one cardiac cycle. In addition, semi-automated measurements have been developed and validated [15,16]. Although the precision of one measurement is presumed to be approximately 100 μ with classical echo devices, this limit is estimated at approximately 30 μ with echo-tracking [15]. By averaging multiple measurements, accuracy can be increased. Which arterial wall should be measured? Reliable IMT measurements can be obtained from the far wall whereas measurements at the near wall should not be considered because they have to be performed at the trailing edge of the ultrasound pulse [12]. As a result, these measurements depend on gain settings and resolution characteristics of the ultrasound device. Wong et al. [17] reported that, as a result, the near wall may represent only 80% of the histological thickness. In a recent study, the IMT difference between the near and far wall was much less [13]. In some large studies, IMT of the near and far wall have been used to increase the number of measurements and the statistical power of the study. This may lead to these IMT values underestimating true IMT, making any comparison with studies using only far wall measurements less appropriate. According to Wong et al. [17], the anticipated underestimation of this averaging procedure is approximately 10%. In a population sample of 250 subjects with an age range between 27 and 82 years, mean carotid IMT was 661 μ [12]. In this population, the average underestimation of the near and far wall averaging procedure would be 66 μ. IMT of the carotid artery is not constant. It continuously changes during the heart cycle. It decreases during systole due to the larger vessel diameter and increases during diastole. It is assumed that the cross-sectional area of the IMT is constant, although compressibility of the intima–media layer due to the higher pressure in systole has not been studied quite so well. Meinders et al. [18] reported differences in IMT compressibility due to spatial inhomogeneities. In a small sample of 44 subjects aged 18–83 years, the change in IMT of the common carotid artery during the heart cycle averaged 25 μ [19]. Assuming that IMT is linearly altered with increasing diameter from diastole to systole, and that, similar to blood pressure, the mean change in IMT diameter during the cardiac cycle is close to one-third of the maximal IMT change, then the average difference between IMT in diastole and mean IMT during the cardiac cycle is estimated to be 8 μ. Although measurement of IMT at a fixed timepoint during the cardiac cycle will decrease variability between data, for an average IMT of 661 μ, the maximal error due to IMT changes during the cardiac cycle is 3.8% and the difference between diastolic and mean IMT is 1.3%. We can conclude that IMT variation during the cardiac cycle is present but, compared to averaging near and far wall and the accuracy limits of the devices, is small and often negligible. In clinical trials, the effects of vasodilating antihypertensive drugs, such as calcium antagonists [4,6,20,21] and angiotensin-converting enzyme inhibitors [22], on carotid IMT have been studied. In these studies, IMT changes per year are quite small. Therefore, it is important to correct for changes in diameter or to calculate cross-sectional intima–media area [14] to distinguish a vasodilation-induced functional decrease in IMT from a real structural decrease in IMT. Recently, a higher IMT was shown to be accompanied by a larger arterial diameter [23]. This demonstrates the complex interaction between IMT and arterial diameter and suggests that progression rate of IMT underestimates the progression rate of the corresponding cross-sectional intima–media area. The compensatory increase in arterial diameter appears limited to an IMT of 1300 μ [23]. The best site of carotid IMT measurements is still a matter of debate. Often, IMT is measured at three different sites [6,11,14]: (i) the common carotid artery (1–2 cm proximal to the flow divider); (ii) the bifurcation and (iii) the internal carotid artery. The Carotid Atherosclerosis Progression Study shows a larger IMT progression rate and a better correlation with cardiovascular risk factors at the internal carotid artery compared to the common carotid artery [11]. Progression rate of the carotid bifurcation was between the internal and common carotid artery, but did not differ statistically from both arterial sites. However, it is not clear whether absolute IMT and/or progression rate of the internal carotid artery and/or bifurcation would better predict cardiovascular events. Some new arterial sites, such as the common femoral artery, have been proposed for IMT measurements. Where the elastic common carotid artery is assumed to represent mainly intimal thickening, IMT of the more muscular common femoral artery represents a change in the intima and/or media layer. The question is whether femoral artery IMT thickening, and thickening of IMT of muscular arteries in general, has the same meaning as carotid artery IMT thickening. This doubt is supported by several studies. There are conflicting data on the relationship between femoral artery IMT as a predictor of coronary atherosclerosis [24,25]. The APSIS study [26] showed that IMT of the common femoral artery has prognostic implications in patients with stable angina pectoris but is different from common carotid artery IMT. In the present issue of the journal, Agewall et al. [27] describe an association between left ventricular hypertrophy and IMT of both the common carotid and the brachial artery in patients with previous myocardial infarction. An association between the common carotid and brachial artery IMT was present in the univariate analysis, but disappeared in the multivariate analysis, showing that the two IMTs are differently influenced and suggesting that they do not have the same meaning. It has been shown repeatedly that the muscular arteries can react to situations by changing the medial thickness. Girerd et al. [28] described an increased IMT at the radial artery in hypertensive subjects due to medial hypertrophy. In a recent study [29], increased IMT of the femoral artery, but not the carotid artery, was also found to be related to higher blood pressure. In the study by Agewall et al. [27], patients with previous myocardial infarction and left ventricular hypertrophy were studied. This suggests a condition of remodelling after myocardial infarction with an activated renin–angiotensin system. Increased angiotensin II levels promote not only cardiac, but also vascular hypertrophy, which can lead to a different effect on the muscular brachial artery than on the elastic common carotid artery. This is supported by the study results showing that, in the univariate analysis, only IMT of the brachial, and not of the carotid artery, is associated with a low ejection fraction, a condition with a triggered renin–angiotensin system. In conclusion, by contrast to the common carotid artery, in more muscular arteries such as the common femoral and brachial arteries, IMT can largely be influenced by an increased medial layer. An increased medial layer has not been shown to be a marker of generalized atherosclerosis. Studies repeatedly suggest that increased IMT of the muscular arteries does not have the same meaning as carotid artery IMT. More studies are required to show the prognostic value of IMT of muscular arteries. In addition, the carotid artery can be accessed more easily, measurements are reproducible and a large amount of data exist [30]. Therefore, for daily clinical practice, it is advisable to stick to IMT measurements of the common carotid artery. There are indications that the internal carotid artery and/or the bifurcation might be more predictive sites than the common carotid artery [11], but this still has to be confirmed in longitudinal studies. Because changes in IMT are generally small over time, determination of IMT progression rates in individual patients will be hard to obtain, mainly because of the accuracy limits of the devices. To avoid misinterpretation of IMT progression rates in longitudinal studies, it is advisable that measurements are performed by the same investigator, at the same arterial site, using the same technique and identical equipment. Images should be centrally calculated by a blinded observer. It would also appear to be necessary to build in control measurements to correct for spontaneous changes in measurements over time [13].