3D critical plaque wall stress is a better predictor of carotid plaque rupture sites than flow shear stress: An in vivo MRI-based 3D FSI study

Abstract

Atherosclerotic plaque rupture leading to stroke is the major cause of long-term disability as well as the third most common cause of mortality. Image-based computational models have been introduced seeking critical mechanical indicators, which may be used for plaque vulnerability assessment. This study extends the previous 2D critical stress concept to 3D by using in vivo magnetic resonance image (MRI) data of human atherosclerotic carotid plaques and 3D fluid-structure interaction (FSI) models to: identify 3D critical plaque wall stress (CPWS) and critical flow shear stress (CFSS) and to investigate their associations with plaque rupture. In vivo MRI data of carotid plaques from 18 patients scheduled for endarterectomy were acquired using histologically validated multicontrast protocols. Of the 18 plaques, histology-confirmed that six had prior rupture (group 1) as evidenced by presence of ulceration. The remaining 12 plaques (group 2) contained no rupture. The 3D multicomponent FSI models were constructed for each plaque to obtain 3D plaque wall stress (PWS) and flow shear stress (FSS) distributions. Three-dimensional CPWS and CFSS, defined as maxima of PWS and FSS from all vulnerable sites, were determined for each plaque to investigate their association with plaque rupture. Slice-based critical PWS and FSS were also calculated for all slices for more detailed analysis and comparison. The mean 3D CPWS of group 1 was 263.44 kPa, which was 100% higher than that from group 2 (132.77, p=0.03984). Five of the six ruptured plaques had 3D CPWS sites, matching the histology-confirmed rupture sites with an 83% agreement. Although the mean 3D CFSS (92.94 dyn/cm(2)) for group 1 was 76% higher than that for group 2 (52.70 dyn/cm(2)), slice-based CFSS showed no significant difference between the two groups. Only two of the six ruptured plaques had 3D CFSS sites matching the histology-confirmed rupture sites with a 33% agreement. CFSS had a good correlation with plaque stenosis severity (R(2)=0.40 with an exponential function fitting 3D CFSS data). This in vivo MRI pilot study using plaques with and without rupture demonstrates that 3D critical plaque wall stress values are more closely associated with atherosclerotic plaque rupture then critical flow shear stresses. Critical wall stress values may become indicators of high risk sites of rupture. Future work with a larger population will establish a possible CPWS-based plaque vulnerability classification.

Fig. 1

A rendered 3D view of a human atherosclerotic carotid plaque with prior rupture verified by the presence of ulceration

Fig. 2

Selected MRI slices, segmented contours, and histological data of the carotid plaque shown in Fig. 1. (a) T1-weighted in vivo MR images (red asterisks denote the lumen); (b) segmented contours showing lumen, outer wall, and atherosclerotic components; and (c) presence of a penetrating ulceration validated by histology (S6).

Fig. 3

Band plots of plaque wall stress (stress-P₁) and FSS for the plaque sample given in Fig. 1 showing CPWS was able to predict the actual rupture site while CFSS failed to do so. (a) A stack view of PWS, (b) a stack view of FSS, and (c) PWS plot on a longitudinal cut showing the bifurcation.

Fig. 4

Comparison of mean slice-CPWS from groups 1 and 2 and from slices within group 1 showing that higher CPWS values may be linked to plaque rupture. (a) Comparison of mean slice-CPWS values between group 1 and group 2 (110.20 ± 78.33 kPa versus 83.40 ± 33.42 kPa; P = 0.00152) and (b) comparison of mean slice-CPWS values between the slices with ulceration (n = 11) and the slices without ulcerations (n = 37) in the rupture group (196.11 ± 114.25 kPa versus 84.65 ± 38.19 kPa, P = 0.00001).

Fig. 5

No statistically significant differences were found from comparison of mean slice-CFSS from groups 1 and 2 and from slices within group 1 with and without ulceration. (a) Comparison of mean slice-CFSS values between the group 1 (n = 48) and group 2 (n = 109), P = 0.16017 and (b) comparison of mean slice-CFSS values between the slices with ulceration (n = 11) and the slices without ulcerations (n = 37), P = 0.26951.

Fig. 6

Correlation of slice-CFSS and slice-CPWS with NWI. (a) slice-CFSS increases exponentially as NWI increases (R2=0.52) and (b) slice-CPWS shows a weak correlation with NWI (r = 0.0161).

Fig. 7

Correlation of 3D CFSS and 3D CPWS with NWI