(D-131) Vorticity as a Predictive Marker for Adverse Cardiac Events in Coronary Artery Disease

Friday, October 13, 2023

3:30 PM – 4:30 PM PDT

Location: Exhibit Hall - Row D - Poster # 131

Presenting Author(s)

Maximilian J. Huber (he/him/his)

Undergraduate Researcher
Duke University
Durham, North Carolina, United States
JJ

Juliet M. Jiang (she/her/hers)

Undergraduate Student
Duke University
WINCHESTER, Massachusetts, United States

Co-Author(s)

Cyrus Tanade, MEng

PhD Candidate |NSF GRFP Fellow
Duke University
Durham, North Carolina, United States

Primary Investigator(s)

AR

Amanda Randles

Alfred Winborne and Victoria Stover Mordecai Assistant Professor of Biomedical Sciences
Duke University, United States

Introduction:: Coronary artery disease (CAD), the leading cause of mortality in the US, arises from the development of fatty plaques that impede blood flow (e.g., lead to stenosis) in the coronary arteries [1]. The clinical standard to gauge the severity of CAD is fractional flow reserve (FFR); the cutoff FFR < 0.8 determines when surgical intervention is necessary [2]. Although effective, FFR has its shortcomings, including the need for invasive measurement and a significant gray zone (GZ) where FFR is largely inconclusive
(0.75 ≤ FFR ≤ 0.85) [3]. The untreated GZ-FFR cases in the range 0.8 ≤ FFR ≤ 0.85 are subject to a substantially higher burden of major adverse cardiovascular events (MACE), which is defined as the combination of the following outcomes:
myocardial infarction, target vessel revascularization, unstable angina, stroke, and death [4].

Existing literature has established that FFR can be computed non-invasively through computational fluid dynamics (CFD) simulations [5, 6, 7]. Furthermore, previous work proposed vorticity as a novel biomarker since it varies significantly near stenotic regions in CAD patients [8]. Vorticity captures the tendency of blood to rotate along the central luminal axis of the artery and
is a measure of locally disturbed flow. Additionally, vortical patterns are understood to impede blood flow in
stenotic arteries [9]. In this analysis, we increase temporal resolution relative to prior studies and calculate vorticity at
every timestep throughout the cardiac cycle and in several key regions around the stenosis to explore its utility for
predicting MACE in untreated GZ-FFR patients.

Materials and Methods::

Our study is based on a retrospective analysis of 19 patients in the untreated GZ-FFR range in whom FFR values were determined by invasive angiography, and follow-up was recorded after 22 months to assess MACE. Four participants experienced complications and were included in the ’GZ-yes’ group, while the rest became the ’GZ-no’ group (Fig. 1a).

Coronary angiograms were used to construct digital twin 3D geometries [10]. Images were captured at the end of diastole and were validated by an experienced cardiologist, who annotated the location of the stenosis. CFD simulations used HARVEY, a massively parallel lattice Boltzmann method hemodynamics solver, which simulated one cardiac cycle of blood flow in the coronary arterial tree for each patient [11]. Vorticity was calculated as the magnitude of curl in the velocity vector field, resulting in scalar measurements for curl at each fluid node included in the analysis (Fig. 1b). The vorticity at each fluid node is calculated as follows (1), given the velocity matrix A = (u, w, v).

∇ΧA = (∂/∂x,∂/∂y,∂/∂z) X (u,w,v). (1)

Vorticity was calculated in this manner in three distinct slices near the stenosis, the central slice was at the minimum luminal diameter (MLD) of the stenosis, and the proximal and distal slices were 2-5cm from the center slice in either direction (Fig. 1c). We also chose to evaluate vorticity at two distinct time frames: averaging vorticity throughout each step in the cardiac cycle, and independently analyzing vorticity at end-diastole timestep since X-ray angiography images were captured then.

Results, Conclusions, and Discussions:: Our analysis shows the early promise of vorticity as a predictive biomarker of MACE in untreated GZ-FFR patients. The difference between the GZ-yes and GZ-no groups was most significant when comparing the maximal vorticity in the proximal slice at the end-diastole timestep; this metric was calculated in all patients and was averaged within the GZ-yes and GZ-no groups (p = 0.02, Fig. 1d). Interestingly, vorticity at the proximal location demonstrated greater significance in differentiating between the GZ-yes and GZ-no groups than in prior analysis, where maximal vorticity was averaged across the three slice locations at end-diastole [8]. This can be explained by the further observation that vorticity tends to be elevated in the MLD and distal slice locations regardless of patient outcome, demonstrating the expected typical occurrence of vortical formations after stenosis (Fig. 1d). On the other hand, proximal vorticity appears to be a much more sensitive indicator of MACE. Perhaps it captures the phenomenon of deleterious entrance effects into the stenotic region reducing the energy of blood flow and contributing to disordered flow, which has been shown to increase plaque development [12].

However, there were limitations in the power of analysis due to the small group in the data set who experienced MACE. Vorticity was elevated in the medial and distal slices regardless of the patient outcome; therefore, no significant conclusions could be reached regarding the impact of vorticity in those locations. The small size of the group also limits insight into the cause of proximal vortical flows, which is a direction for future work. Further work should also evaluate other recorded hemodynamic parameters, such as velocity, to determine if elevated proximal vorticity results in a net velocity reduction along the artery.

This work indicates that vorticity should be further validated as a biomarker to inform surgical intervention in patients with untreated GZ-FFR, as it appears to offer significant clinical utility. It also points towards a promising future for CFD simulations, adding novel insight to elevate the standard of care and addressing an unmet need in treating the most deadly disease in the United States.

Acknowledgements (Optional): :

References (Optional): :

[1] J. C. Brown, T. E. Gerhardt, and E. Kwon, “Risk Factors for Coronary Artery Disease,” in StatPearls, Treasure Island (FL): StatPearls Publishing, 2023. Accessed: Jul. 16, 2023. [Online]. Available: https://www.ncbi.nlm.nih.gov/books/NBK554410/
[2] I. Y. Elgendy, C. R. Conti, and A. A. Bavry, “Fractional Flow Reserve: An Updated Review,” Clin Cardiol, vol. 37, no. 6, pp. 371–380, Mar. 2014. doi: 10.1002/clc.22273.
[3] “Fractional Flow Reserve–Guided Revascularization: Practical Implications of a Diagnostic Gray Zone and Measurement Variabil-
ity on Clinical Decisions — JACC: Cardiovascular Interventions.”
https://www.jacc.org/doi/abs/10.1016/j.jcin.2012.10.014 (accessed Jul.16, 2023).
[4] R. Abad Romero et al., “Myocardial infarction with non-obstructive coronary arteries (MINOCA): predictors of long-term outcome,” European Heart Journal, vol. 43, no. Supplement2, p. ehac544.1460, Oct. 2022, doi: 10.1093/eurheartj/ehac544.1460.
[5] I. Pantos and D. Katritsis, “Fractional Flow Reserve Derived from Coronary Imaging and Computational Fluid Dynamics,” Interv Cardiol, vol. 9, no. 3, pp. 145–150, Aug. 2014, doi: 10.15420/icr.2014.9.3.145.
[6] X. Xie, M. Zheng, D. Wen, Y. Li, and S. Xie, “A new CFD based non-invasive method for functional diagnosis of coronary stenosis,” BioMedical Engineering OnLine, vol. 17, no. 1, p. 36, Mar. 2018, doi: 10.1186/s12938-018-0468-6.
[7] K. M. Chinnaiyan et al., “Rationale, design, and goals of the HeartFlow assessing diagnostic value of non-invasive FFRCT in Coronary Care (ADVANCE) registry,” Journal of Cardiovascular Computed Tomography, vol. 11, no. 1, pp. 62–67, Jan. 2017, doi: 10.1016/j.jcct.2016.12.002.
[8] M. Vardhan et al., “Evaluation of intracoronary hemodynamics identifies perturbations in vorticity,”
Frontiers in Systems Biology, vol. 2, 2023, Accessed: Jul. 16, 2023. [Online]. Available: https://www.frontiersin.org/articles/10.3389/fsysb.2022.930396
[9] M. Amir et al., “Analysing Spatio-temporal flow hemodynamics in an artery manifesting stenosis,” International Journal of Mechanical Sciences, vol. 218, p. 107072, Mar. 2022, doi: 10.1016/j.ijmecsci.2022.107072.
[10] S. J. Chen and J. D. Carroll, “3-D reconstruction of coronary arterial tree to optimize angiographic visualization,” IEEE Transactions on Medical Imaging, vol. 19, no. 4, pp. 318–336, Apr. 2000, doi: 10.1109/42.848183.

[11] A. Randles, E. W. Draeger, and P. E. Bailey, “Massively parallel simulations of hemodynamics in the primary large arteries of the human vasculature,” J Comput Sci, vol. 9, pp. 70–75, Jul. 2015, doi:10.1016/j.jocs.2015.04.003.
[12] “Inflammation and enhanced atherogenesis in the carotid artery with altered blood flow in an atherosclerosis-resistant mouse strain - Zhao - 2021 - Physiological Reports - Wiley Online Library.” https://physoc.onlinelibrary.wiley.com/doi/10.14814/phy2.14829 (accessed Jul. 16, 2023).