Background
The central mechanism of atherosclerosis is chronic inflammation in the presence of damaged vascular endothelium and lipid-laden foamy macrophages derived from infiltration of monocytes into the arterial wall. This mechanism can lead to coronary stenosis and thrombotic obstruction after disruption of the resulting atherosclerotic plaque [
1]. Accumulation of leukocytes and lipids, and proliferation of smooth muscle cells, cell death, and fibrosis occur on the damaged endothelium [
2]. Although the arterial wall is exposed to risk factors, such as systemic hypertension, hypercholesterolaemia, and diabetes, atherosclerotic plaques develop preferentially at specific areas [
3]. In patients with acute coronary syndrome (ACS), the distribution of ruptured coronary artery plaques in the lumen is significantly more eccentric than that of non-ruptured plaques. This finding suggests that blood flow influences the location of ruptured plaques and may even contribute to plaque rupture [
4]. The relationship between the spatial distribution and the phenotype of plaques under conditions where blood flow influences atherosclerosis in stable patients has not been fully elucidated. In this study, we used grey-scale intravascular ultrasound (IVUS) to identify spatial plaque distribution, and virtual histology (VH)-IVUS to evaluate the plaque phenotype in 30 consecutive patients who underwent elective percutaneous coronary intervention (PCI), in an attempt to clarify the association between geographical predisposition and plaque phenotype.
Discussion
The major findings of this study are as follows. (1) Eccentric plaques were more frequently distributed towards the myocardial side than towards the lateral and epicardial sides of the coronary artery. (2) A significant difference was observed in the diameters of the eccentric plaque vessels between the distributed sides. (3) The difference in the plaque component between the distributed sides was also significant. TCFAs were more frequently observed in myocardial side plaques than in lateral or epicardial side plaques.
Coronary arteries are continually subjected to mechanical force, such as tensile or compressive stress and shear stress generated by the heartbeat and pulsatile blood flow during each cardiac cycle [
11,
12]. Among the biomechanical forces, flow generates tangential drag force and resultant shear stress. The magnitude of shear stress is determined by changes in luminal geometry, blood flow velocity, and plasma viscosity [
13]. Blood flow is disturbed by vessel curvature; it is fast in the outer curvature and slow in the inner curvature. Shear stress is high in the outer curvature and low in the inner curvature [
13‐
15]. Endothelial cells sense shear stress and alter their shape and phenotype [
16]. The shear stress is controlled by adapting vessel size to suit the blood flow in response to sustained changes [
17]. Hypothetically, if coronary arteries were classified geometrically, myocardial, epicardial, and lateral sides would be exposed to low, high, and intermediate shear stress, respectively. Although shear stress may change over time as plaque progression into the lumen changes coronary flow, we found that eccentric plaques were more frequently distributed towards the myocardial side than towards the epicardial side or lateral side, which is consistent with the hypothesis mentioned above.
Vascular adaptation by shear stress allows the arterial tree to deviate from a straight-tube geometry to another morphology. This phenomenon permits the shear stress to remain unchanged, which provides the predilection site for eccentric plaque development [
18]. Human autopsy data showed compensatory enlargement of human coronary arteries in relation to plaque area, and lumen stenosis was delayed until the lesion occupied 40% of the internal elastic lamina, which is termed Glagov’s phenomenon [
19]. In the present study, the plaque area in the myocardial and epicardial sides was significantly larger than that in the lateral side, although no significant difference was found in the lumen area between the distributed sides; this finding indicates compensatory enlargement of the myocardial and epicardial vessels. We also confirmed that the minimum vessel diameter was significantly larger on the epicardial side than on the lateral side, and that the maximum lumen diameter was significantly larger on the myocardial side than on the lateral side. Although the difference was not statistically significant, the average vessel diameter was numerically larger on the myocardial side than on the lateral side (4.43 ± 0.04 mm vs. 4.15 ± 0.05 mm,
p = 0.108), and numerically larger on the epicardial side than on the lateral side (4.48 ± 0.07 mm vs. 4.15 ± 0.05 mm,
p = 0.0814). The average plaque area was approximately 40% (44.83 ± 0.75%), and the compensatory vascular remodelling was associated with geographical predisposition.
Using VH-IVUS imaging, a spatial relationship between low shear stress and the necrotic core was observed in early plaques (plaque burden <40%) [
20], while increases in the necrotic core percentage occurring at the site were typically affected by low shear stress [
21]. In serial observations of endothelial shear stress and plaque composition, low-stress segments had greater plaque and necrotic core progression compared with intermediate-stress coronary segments, and high-stress segments had greater necrotic core and calcium progression [
22]. In the present study, analysis of plaque composition by VH-IVUS revealed that lateral and epicardial plaques contained significantly more fibrous plaque component than myocardial plaques. Myocardial side plaques contained less fibrous component than the lateral and epicardial side plaques, whereas myocardial side plaques contained more fibro-fatty area than lateral plaques. Myocardial and lateral side plaques contained more necrotic core component than the epicardial side plaques, and epicardial plaques contained more calcium than the lateral plaque. The actual proportion of each plaque component correlated well with assumed shear stress being high on the epicardial side, intermediate on the lateral side, and low on myocardial side (Table
3).
In a previous study using integrated backscatter IVUS [
23], Sato et al. reported that in plaques with moderate stenosis in non-branching lesions, lipid pools clustered in the inner curvature and fibrous tissue clustered in the outer curvature. In accordance with their findings, we also found that fibro-fatty and necrotic contents identified by VH-IVUS were more often seen in myocardial side plaque. Although they studied both eccentric and concentric plaques, whereas we selected only the eccentric plaques for analysis, different imaging modalities specifically useful for plaque content characterisation confirmed similar results.
Longitudinal studies in porcine models have shown that TCFAs, which develop more frequently in the coronary regions, are exposed to low shear stress throughout their evolution [
24,
25]. Autopsy studies have shown that atherosclerotic lesions are provoked by TCFA rupture, which can lead to thrombosis, ACS, and sudden cardiac death [
26,
27]. In vitro studies using the finite element method have demonstrated that the shear stress of the vascular lumen is an important determinant of coronary plaque vulnerability and plaque rupture [
28,
29]. Fukumoto et al. demonstrated that localised elevation of blood pressure and shear stress are associated with coronary plaque rupture in the proximal or top portion of the plaque in ACS patients [
30]. The shear stress concentration is frequently correlated with the plaque rupture site. Plaque rupture may heal without any symptoms or lead to mural thrombosis with subsequent asymptomatic healing [
31,
32].
Although the precise mechanisms that promote the focal formation of rupture-prone coronary plaques in vivo remain to be elucidated, we found that eccentric TCFAs were clustered towards the myocardial side. We only analysed eccentric plaques, which may be predisposing to future coronary events [
4]. The relationship between rupture-prone TCFAs and subsequent thrombus formation or clinical events is still unknown [
31], as is whether TCFA-induced plaque ruptures lead to lumen stenosis. Although it is also still unclear whether TCFA clusters towards the myocardial side actually rupture and lead to clinical symptoms or lumen stenosis, the method for the geographical classification of coronary plaques by using IVUS in this study is simple and applicable in clinical settings, and can be utilized to characterise the complex profile of atherosclerotic plaque.
Study limitations
The limitations of this study are as follows. First, the sample size was small; only 30 coronary arteries in 30 patients were analysed. Second, all the patients were in stable condition, and their plaque phenotype may have been different from that of unstable patients. Third, we included right arteries, in which atherosclerotic change may differ from that in left coronary arteries [
33]. Fourth, we did not calculate the absolute value of inter-observer variability in identifying the distribution of plaque, although this does not invalidate the findings because discordance in the image reading was rare. Fifth, this study was designed as an observational study, and the clinical importance of geographical predisposition should be assessed prospectively.
Acknowledgements
We thank Michinori Iwasaki for valuable technical support.