Carotid artery intima-media thickness is closely related to impaired left ventricular function in patients with coronary artery disease: a single-centre, blinded, non-randomized study

The measurement of carotid intima-media thickness (IMT) has long been regarded as a method which can be used to evaluate the presence of generalized atherosclerotic arterial disease [1, 2]. In some reports carotid IMT values have been related to left ventricular (LV) hypertrophy and function [3, 4]. One study of patients with carotid disease showed a possible prediction of minimal cardiovascular resistance in patients with probable CAD [5]. LV function can be assessed with echocardiography using an ejection fraction and wall motion score index, but this method is highly operator dependent. A more reliable method that could detect early LV dysfunction would therefore be of great clinical value. Two-dimensional speckle-tracking echocardiography (2D-STE) is a semi-automated quantitative technique for the assessment of strain in the LV and is a well validated method [6, 7].

Strain is an intrinsic mechanical property which measures myocardial systolic function more accurately and has the ability to detect early pathological changes better than conventional cavity-based echocardiographic measurements [8‐10].

The LV wall is a 3-dimensional continuum of myocardial fibers with an inner right-handed helix (endocardium), a middle circular layer and an outer left-handed helix (epicardium). The endocardial layer is the most susceptible to ischemic injury. Layer-specific 2D-STE allows the assessment of LV function in all 3 layers.

Carotid IMT may be used to assess the risk for CAD [1, 2]. Ischemia due to CAD is one of the causes of decreased LV function and there is therefore a possibility that carotid IMT may be used to assess LV function in these patients. The aim of this study was to investigate a possible correlation between IMT and LV function, assessed using strain echocardiography, in patients with CAD.

Thirty- one patients took part in the study. Two patients were excluded due to poor echocardiographic image quality. Strain analyses were therefore performed in 29 patients.

The age of the patients ranged from 53 to 75 years. There were 81% males in the study. Baseline characteristics of the patients are shown in Table 1.

A total of 20 (65%) patients had significant CAD (≥50% stenosis). Eight of these patients had an occlusion of a major coronary artery) and 11 (35%) had a non-significant CAD (<50% stenosis). There was a significant difference in endocardial and mid-myocardial LVGLS between the patients with significant coronary stenosis and the patients with non-significant coronary stenosis. Layer-specific LVGLS compared to the coronary angiography findings are shown in Table 2.

Left ventricular Ejection Fraction (LVEF) was: mean ± SD; 61% ± 6% for the total group of patients. Patients with normal endocardial strain had EF values of: mean ± SD; 62% ± 6.0% and patients with reduced endocardial strain mean ± SD; 58% ± 4.8%. This difference was not significant (p = 0.091).Figures 4a-d shows ROC analyses of layer specific strain parameters and the ability of IMT measurements to identify patients with significant CAD. Maximum carotid IMT ranged between 1 mm and 2.88 mm for the whole group of patients including carotid plaques in the IMT measurements. For the group with normal endocardial strain values, defined by the cut-off strain value of -16.7, carotid IMT was 1.2 mm ± 0.2 (mean ± SD) and for the group with reduced endocardial strain values the carotid IMT value was 1.7 mm ± 0.5 (mean ± SD). Independent samples t-test comparing the two groups was statistically significant (p = 0.006). There was significant correlation between endocardial LVGLS and maximum carotid IMT (p = 0.006). The results of linear regression analyses which incorporated adjustment for known risk factors of atherosclerosis (hypertension, smoking, hyperlipidemia, diabetes and BMI) were made and showed significant results (p = 0.02).

In this study we found a significant correlation between increased carotid IMT and LV dysfunction. We also found associations between significant CAD assessed by coronary angiography and reduced endocardial and mid-myocardial function assessed by 2D speckle-tracking strain echocardiography.

The cut-off value for endocardial strain was approximately -16.7% whereas normal strain values are approximately -20% [13]. We had no gold standard with regard to normal and abnormal 2D-STE values which could be used for the ROC curves comparing carotid IMT versus 2D-STE. We therefore calculated a cut-off value for pathological endocardial strain measurements. We found affection of LV function in the endocardial and midmyocardial layers but not in the epicardial layer. This finding is in accordance with the fact that, reduced myocardial function usually starts in the endocardial layer and this layer is the most susceptible layer for ischemic injury [14].

There are few studies investigating the association between carotid disease and early left ventricular injury but there are some findings which correspond to the present study. The correlation between LV dysfunction and increased carotid IMT was found in one study of individuals free of clinical CAD, where greater carotid IMT was associated with changes of myocardial strain, measured by tagged MRI [15]. Mizuguchi et al. found that LV relaxation measured by 2D-STE was associated with carotid arterial atherosclerosis in particular related to sclerosis, measured by carotid arterial stiffness. They did not find this correlation for carotid IMT and LV strain [16].

The assessment of LV function is challenging due to its very complex deformation pattern. All imaging modalities have shortcomings in this regard. However, although myocardial strain measures are only a component of LV function, it has been demonstrated to be an accurate and reproducible measure in several studies [6, 7].

All patients in our study had symptoms of CAD or risk factors for atherosclerosis. It is known that structural and functional changes in the carotid arteries occur with normal aging. However, the presence of cardiovascular risk factors accelerate these changes [17‐19].

It is well known that carotid intima-media thickness and plaques have different pathophysiological mechanisms. However, both carotid IMT and plaques are well-known indicators of increased risk for cerebro- and cardiovascular disease [20, 21]. We therefore included the diameter of the small number of plaques when detected in the area of interest.

In a study of patients with hypercholesterolemia and preserved LVEF, 1 year treatment with a statin improved regional LV systolic and diastolic function measured by strain [22]. Carotid stiffness was lower after statin therapy, but there was no difference in carotid IMT before and after treatment with a statin. This is probably due to the fact that the annual rate of IMT progression is lower than the resolution of carotid ultrasound (≈0.3 mm). The presence of increased IMT may lead to detection of very early cardiac dysfunction which may benefit from cardiovascular risk management.

The results of this study support the few previous studies which have examined a possible association between carotid IMT and LV strain. B-mode ultrasound of the carotid artery is a non-invasive, inexpensive and widely used examination and may therefore be an additional appropriate tool for the assessment of possible CAD. An ultrasound examination of the carotid arteries is mandatory in all cases with cerebrovascular disease. A more detailed examination which includes IMT measurements in patients with cardiovascular risk factors may therefore be of value when estimating the risk of subclinical coronary artery disease.

In this study increased carotid IMT values were associated with decreased left ventricular function. These findings support the use of carotid IMT measurements to predict the risk of coronary heart disease.