The underlying cause of most cardiovascular diseases is atherosclerosis, an inflammatory-driven disease associated with lipid accumulation in the vessel wall, resulting in plaque formation. The destabilization and rupture of an atherosclerotic plaque remain a leading cause of death. In recent decades, an increasing amount of research has focused on the contribution of local risk factors precipitating future plaque rupture and subsequent cardiovascular events [1‐3].

One of these local risk factors is the composition of the plaque. The presence of a lipid pool inside the plaque has been strongly associated with plaque vulnerability, destabilization, and clinical events [3, 4]. Such lipid pools in the vessel wall can currently be detected using the invasive imaging technique near-infrared spectroscopy (NIRS). A combined catheter of NIRS with intravascular ultrasound (IVUS) has been validated for the detection of the local lipid content of coronary plaques [5]. Previous studies using NIRS have shown that the presence of these lipid-rich plaques (LRPs) is independently associated with future major adverse cardiovascular events [3, 4].

A second local factor that is strongly associated with atherosclerosis is wall shear stress (WSS), the frictional force of the bloodstream exerted on the endothelial cells of the vessel wall. Low WSS causes endothelial dysfunction and plays an important role in the development of atherosclerosis [6]. In more advanced stages of atherosclerosis, if plaques intrude into the lumen, WSS will increase. High WSS has been related to apoptosis of smooth muscle cells and consequently to plaque vulnerability and potentially plaque rupture [7]. Recently, high WSS has been proven to be of added value in the prediction of myocardial infarction [8].

Although both high lipid content detected by NIRS and high WSS are associated with plaque destabilization and cardiovascular events, knowledge on co-localization of these two factors is still lacking. We hypothesized that high WSS plays a role in the destabilization of LRPs and thereby contributes to an event. Therefore, in this present study, the co-localization of LRPs and high WSS is investigated in non-culprit segments of patients presenting with an acute coronary syndrome (ACS) as a first step to understand the potential interaction between high WSS and LRPs in plaque destabilization and future events.

Between March 2016 and March 2018, a total of 53 patients were enrolled. Four patients withdrew study consent between the baseline procedure and CCTA. There were no major adverse events related to the study procedure. In one patient, it was not possible to match the CCTA-derived centerline with the NIRS-IVUS due to motion artifacts. In one vessel, no NIRS data was acquired during the NIRS-IVUS pullback, and in twelve vessels, less than 1% of the chemogram contained NIRS(+) data. These were all excluded from the analysis. Consequently, 38 vessels from 37 patients were included in the current analysis. Baseline characteristics are listed in Table 1. The mean age of the patients was 62 ± 9 years, 91.9% was male, 24.3% of the patients had a history of previous PCI, and 40.5% of the patients were on statin therapy at the time of inclusion in the study. The studied non-culprit segment was located in the left anterior descending artery in 39% of the cases, in the left circumflex in 29%, and in the right coronary artery in 32% of the cases.

The present study shows for the first time that LRPs co-localize with exposure to high levels of TAWSS. Additionally, for the most diseased regions in the vessel, as identified by the MaxLCBI_4mm, a dose-dependent relation was found between LCBI and exposure to high TAWSS.

LRPs, as detected by NIRS, have been known to be a risk factor for future plaque rupture and major cardiovascular events [3, 4]. Since not all LRPs rupture and result in an event, for clinical decision making and invasive treatment strategies, it is of crucial importance to understand the pathophysiology of LRP in its evolution to high-risk vulnerable plaques. Since plaque destabilization has been proven to be associated with high TAWSS [8], as a first step in unravelling LRP destabilization, we investigated the exposure of LRP to high TAWSS. Therefore, we used detailed invasive imaging with NIRS-IVUS and computational fluid dynamics in non-culprit coronary arteries of acute coronary syndrome patients. In order to discuss the relationship between LRP and high TAWSS, it is essential to understand the interplay between lumen geometry, TAWSS distribution, and the pathophysiology of atherosclerosis. The magnitude of TAWSS is affected by both the flow and the lumen area, such that vessel segments with a smaller lumen but with the same flow are exposed to a higher TAWSS. Low TAWSS, mostly observed at inner curves and near side branches, is associated with plaque initiation and progression [15]. If plaque growth leads to minimal local lumen narrowing, this can already result in a local increase in TAWSS. This increased TAWSS is sensed by the endothelial cells with mechanosensors [16] and initiates compensating outward remodeling of the vessel wall, thereby preserving the lumen area and the TAWSS in the earlier stages of the disease [17]. However, Glagov et al. have shown that if cross-sectional plaque burden (plaque area/vessel area ∗ 100) becomes larger than 40%, lumen narrowing can no longer be prevented, and TAWSS subsequently remains high in these regions [18]. The relative plaque area used in this study is a local surrogate for the cross-sectional plaque burden. The higher relative plaque area found per subsequent TAWSS tertile follows the earlier observations by Glagov that cross-sectional plaque burden and lumen narrowing are linked. Therefore, the observed frequent exposure of NIRS(+) sectors to high TAWSS may be explained by the observed higher relative plaque area of these sectors compared to the NIRS(−) sectors.

Although the higher relative plaque area observed for NIRS(+) sectors might be an obvious explanation for the association between LRP and future events, previous prospective NIRS studies have shown that these relationships were independent of plaque burden, minimal lumen area, or clinical characteristics [3, 4]. Recently, a multimodality imaging study confirmed that segments with higher lipid content as detected by NIRS are colocalized with thin-cap fibroatheromas as visualized with optical coherence tomography [19]. Our finding that the majority of regions with high lipid content colocalize with high TAWSS could therefore be of added value in understanding why LRPs are related to cap thinning and future cardiac events. One of the causes of a coronary event is the rupture of the fibrous cap covering an LRP. High TAWSS is suggested to play an essential role in the thinning of the fibrous cap through apoptosis of smooth muscle cells, thereby increasing the plaque vulnerability and thereby the likelihood of rupture [20]. This hypothesis is supported by previous data of our research group, showing that indeed, plaque ruptures are located in regions with elevated TAWSS [21]. To further support this hypothesis, a clinical study by Kumar et al. demonstrated that higher WSS upstream of coronary lesions is predictive for myocardial infarctions, also implying this relationship between high WSS and plaque rupture [8]. The next step in research would be to combine these findings of colocalization of LRPs and high TAWSS and LRPs and thin-cap fibroatheromas to get more insight into the relationship between high TAWSS exerted on LRPs and the development of a cardiac event. Potentially, this could be done in a retrospective study using previously collected NIRS and event data such as the LRP study [4].

A major advantage of this study was the use of the combined NIRS-IVUS catheter. The dual sensor catheter enabled us to simultaneously and precisely detect the NIRS signal, TAWSS, and wall thickness at a high resolution at the same location. This was due to the following three factors: firstly, the 3D model was based on the coronary lumen derived from the IVUS images; secondly, we could precisely map the NIRS signal on top of IVUS images; and thirdly, no matching step was needed to add the NIRS signal to the model.

In the present study, we used relative TAWSS thresholds based on vessel-specific tertile distribution. This method is in contrast to other studies using absolute TAWSS thresholds [1, 22]. The differences in approach are mainly driven by the different methodologies used to determine TAWSS. Firstly, we used 3D models of all individual coronary arteries, which included all the large side branches of each vessel, whereas other studies used simplified geometrical models without side branches. Although the inclusion of side branches makes the computational model more complex, the side branches are vital to include since they affect the absolute TAWSS values of the main branch. Also, given their unique disturbed flow patterns, bifurcation regions are more prone to plaque formation and should be taken into account in the analysis [23]. Secondly, we used Doppler-derived velocity measurements at multiple locations throughout the artery to obtain patient-specific boundary conditions for the CFD simulations, whereas other studies made more general assumptions regarding temporal flow patterns and flow magnitude. Both differences in the approach have a major influence on absolute TAWSS. In our individually tailored model used here, we observed a very large range in TAWSS among patients. When using absolute thresholds, some patients might present with only sectors labeled as low or high TAWSS. To answer if LRP vs non-LRPs within a patient have a different exposure to high TAWSS, absolute thresholds are not suitable when using our current methodology. Furthermore, we excluded the vessels with no LRPs (LCBI < 10, this is 1% NIRS(+)) to allow the comparison of LRPs and non-LRP coinciding in the same vessel.

Firstly, since we aimed at analyzing plaque regions (> 0.5 mm), we excluded all regions with no reliable visualization of EEM, thus also regions with extensive calcifications that block the ultrasound signal. In general, calcified regions are more complex and have thicker plaques and LRPs have shown to colocalize to some extent with extensive calcifications [24]. Therefore, excluding the calcified regions might have led to an overall underestimation of the colocalization of high TAWSS and LRP. If we would assume that calcified regions are most often located in plaques > 0.5 mm, NIRS(+) plaque sectors were even more often exposed to high TAWSS (p < 0.05) (see supplement table). Secondly, in the CFD simulations, we assumed rigid walls and the absence of motion of coronary arteries because of cardiac contraction. However, their effect on time-averages TAWSS quantities has been demonstrated to be minor [25].

In this study, we used detailed invasive imaging and computational fluid dynamics to demonstrate that lipid-rich plaques, as detected by NIRS, are often co-localized with high TAWSS. The co-localization of these two features could be of value in understanding why NIRS(+) plaques are more prone to rupture, leading to clinical events. Future studies are needed to demonstrate the influence of high TAWSS on further lipid-rich plaque development and destabilization.